I don't see how the article being a few years old changes the fact that any metric/analytic that is created by an individual or group of individuals to measure players is going to be completely subjective. Or that the randomness and free-flowing nature of NBA play makes it difficult to measure compared to other sports.
Stats are completely subjective? Okay. I guess you can rely on "alternative stats."
Yes. Because it's weighted depending on what the creator of the metric feels is more valuable or not.
Not really. That's actually a criticism I have of PER and WS/48. That's a reason why I like RAPM-derived stats (like BPM and RPM) though they are all limited in their own right.
Well, here's one area where I agree with Batguano. Every advanced statistic is subjective. Basic stats like FGM are not subjective.
But every single advanced statistic is subjective because advanced stats try to measure things that cannot be measured. Things like efficiency. Or defense.
So that means someone has to decide which variables, constants, multipliers, and so on should or should not be included in the advanced statistic. And then, when they create their initial formula, they have to run it and eyeball it and see if the results make sense to THEM, and if not, tweak the formula until the result set mirrors their own personal opinions about efficiency or defense.
That's been my whole point to him. If he's making the argument that things like "efficiency" are subjective, then why is he so hell bent on trying to argue who is "best". It's no different than Hollinger trying to argue who is most efficient. Doesn't matter whether you use an advanced statistic or observations or maybe neither or you just pick your favorite player and put him there, it's all ... subjective.
So while the effort to quote posts is admirable, it's fruitless. It doesn't mean anything. It doesn't MAKE a player the best anymore than Hollinger's PER makes the player with the higher score necessarily more efficient.
Yeah, right. You really think he would've gone back to Cleveland with no Kyrie and Kevin Love?
No, hence me also saying he saw a great opportunity to do so. I also don't think he would have come to, Houston or Minnesota for example, with the same roster. He wanted to bring a title to his hometown.
Batguano wrote:
It's widely known that LeBron pushed for the Wiggins for Love trade before he would sign on. You're one of those people we mentioned before that bought into the whole BS "I'm coming home" narrative.
I don't think wanting to win one for your home city is BS. It makes no sense to go back to win for your home city when the team is garbage and likely won't win.
But since we are on the team switching/loyalty topic, why does a player get bonus points for staying with arguably the best franchise in the league?
Agree with your latter point. If you're going to give points to players who "toughed it out" then you can't exactly give a player points for re-signing with a stacked team.
Plus, to your point, LBJ went to Cleveland who had come off a 33-win season and all they really added was Kevin Love. And they went from that to immediate contenders in the NBA Finals and Kevin Love didn't even really play!
About part one i think your totally off. What impresses players is everything, not only degree of difficulty, but persistence, consistency and resilience. A hunger to win, never shrinking from the moment, and playing great with anger. If they were only impressed with highlights Vince Carter would be they're guy.
Vince Carter was their guy at one point. Other players love hot dogs as much as the fans do.
Goldenwest wrote:
Kobe was not coasting on reputation. Maybe you can make an argument for 2012 but not for the other years. And even for 2012 he could still lock anyone down pre-Achilles.
Not only can I make the argument, people have already done it. Kobe was an average to mediocre defender even in his prime. He could play elite defense when he wanted to. He usually didn't want to. Here's an example of one of the advanced metrics from 2006:
Steve Freaking Nash rated as a better defender than Kobe.
As I said, most of us knew this at the time. It was an annual ritual on this forum to chuckle at Kobe's automatic election to the all-defense team.
Goldenwest wrote:
you shouldn't blame me cause I'm not going to waste my time countering your silly assertion that Kobe's stats and accomplishments are 'not comparable to Lebron?'. Not sure if you're being facetious or you actually believe that bullsh**, cause that's what it is.
With an analysis you could actually show Kobe's stats and accomplishments to be more impressive which has been somewhat discussed in this thread
Oh, boo hoo. Unless you've had your head in the sand for the last 10 years, you already know that Lebron is one of the greatest statistical players ever. You can pick pretty much whichever of the measures you want -- PER, WS/48, RPM, etc. -- and they all tell you the same thing. Lebron just kills Kobe on stats.
- when the players are talking about the best they talk about Kobe and MJ, they're not talking about Vince Carter and his highlight reel. Don't mix the two together. I laugh when the sports radio host tries to stick Lebron in front but the players resist and prefer Kobe.
-Kobe was always a lock down defender taking on the most prolific scorer in crunch time. He was never an average defender pre-2012. So you know more than the voters? And anyway 'average to mediocre'?
-like i said statistics only tell half the story. and your talking as if there's a big difference between Kobe and Lebron, there isn't. Lebron has an edge in FG% only because he's bigger and thus plays much closer to the rim. But double teams or facing a strong defender? Stats don't tell you about that. That's why watching the games are so important. I've been saying a million times, the stats are not normalized for opponent defensive intensity, teammates reactions or when in the game points are scored. These are big omission in the stats. People like you really pay way too much attention to stats and what your doing is misusing the data to support an opinion without giving much credence to the actual play. Many times people that haven't played much do this. Maybe You should get your own head out of the stat sheet.
Well, here's one area where I agree with Batguano. Every advanced statistic is subjective. Basic stats like FGM are not subjective.
But every single advanced statistic is subjective because advanced stats try to measure things that cannot be measured. Things like efficiency. Or defense.
So that means someone has to decide which variables, constants, multipliers, and so on should or should not be included in the advanced statistic. And then, when they create their initial formula, they have to run it and eyeball it and see if the results make sense to THEM, and if not, tweak the formula until the result set mirrors their own personal opinions about efficiency or defense.
That's been my whole point to him. If he's making the argument that things like "efficiency" are subjective, then why is he so hell bent on trying to argue who is "best". It's no different than Hollinger trying to argue who is most efficient. Doesn't matter whether you use an advanced statistic or observations or maybe neither or you just pick your favorite player and put him there, it's all ... subjective.
So while the effort to quote posts is admirable, it's fruitless. It doesn't mean anything. It doesn't MAKE a player the best anymore than Hollinger's PER makes the player with the higher score necessarily more efficient.
That's not actually true, though... which is my point. RAPM is just based on a simple premise: if a player is good, then his team will outscore the other team more often than when he's not there. The only tricky part is what to include as the prior, and for RAPM, using up to the last three years' past history (or just a "rookies are bad" in general) as the prior information.
The same thing is nearly true of BPM. They just took a bunch of different box score stats, figured out which ones could predict RAPM the best, and used those ones. No one hand chose any variables, constants. The only thing hand chosen is the multiplier for team record (which I've been critical of).
Not every stat is hand crafted like WS/48 and PER (and you'll note I call them bad stats... especially PER). That doesn't mean RAPM or BPM are flawless, not by a long shot. But your specific criticism doesn't actually apply to them.
Not only can I make the argument, people have already done it. Kobe was an average to mediocre defender even in his prime. He could play elite defense when he wanted to. He usually didn't want to. Here's an example of one of the advanced metrics from 2006:
Steve Freaking Nash rated as a better defender than Kobe.
As I said, most of us knew this at the time. It was an annual ritual on this forum to chuckle at Kobe's automatic election to the all-defense team.
Something interesting about Kobe's defensive RAPM which can play either way in your debate about Kobe's defense:
Kobe's defensive RAPM...
In 2005: -1.1
In 2006: -1.2
In 2007: -0.5
In 2008: +0.5
In 2009: +0.4
In 2010: +1.3
In 2011: -0.9
In 2012: +0.2
In 2013: -0.5
What do the positive seasons have in common? They were all seasons where the amount of playoff minutes he played were disproportionately high ('08-'10 due to long playoff runs, '12 because lockout + 7-game 1st round series). Even as he got older. All of his negative seasons? Ones where it was mostly regular season.
The takeaway that's supported both by the fabled eye test and ancillary stats like DBPM is that Kobe turned up his defense in the playoffs. In the context of your debate, it could go either way. On the one hand, the evidence is overwhelmingly clear that most of Kobe's all-defensive awards were not merited. On the flip side, they also suggest he was a lot better as a defender when he was engaged.
I ran a multiple linear regression just to see if the intuition holds true, and indeed it does. The R^2 is even an impressively high 0.73, although this is not rigorous by any means. So I guess the point here is that how good Kobe was on defense largely depends on whether you distinguish between regular season and postseason.
Something interesting about Kobe's defensive RAPM which can play either way in your debate about Kobe's defense:
Kobe's defensive RAPM...
In 2005: -1.1
In 2006: -1.2
In 2007: -0.5
In 2008: +0.5
In 2009: +0.4
In 2010: +1.3
In 2011: -0.9
In 2012: +0.2
In 2013: -0.5
What do the positive seasons have in common? They were all seasons where the amount of playoff minutes he played were disproportionately high ('08-'10 due to long playoff runs, '12 because lockout + 7-game 1st round series). Even as he got older. All of his negative seasons? Ones where it was mostly regular season.
The takeaway that's supported both by the fabled eye test and ancillary stats like DBPM is that Kobe turned up his defense in the playoffs. In the context of your debate, it could go either way. On the one hand, the evidence is overwhelmingly clear that most of Kobe's all-defensive awards were not merited. On the flip side, they also suggest he was a lot better as a defender when he was engaged.
I ran a multiple linear regression just to see if the intuition holds true, and indeed it does. The R^2 is even an impressively high 0.73, although this is not rigorous by any means. So I guess the point here is that how good Kobe was on defense largely depends on whether you distinguish between regular season and postseason.
It also generally correlates with quality of teammates, which could be interpreted a couple different ways. As for your point, I don't know whether the stats include the playoffs. However, I noticed that he didn't have a separate stat for the playoffs, so maybe they do.
-Kobe was always a lock down defender taking on the most prolific scorer in crunch time. He was never an average defender pre-2012. So you know more than the voters? And anyway 'average to mediocre'?
Bull. Kobe was not a lock down defender, and he often was not matched up with the other team's star scorer. We've been through this in many other threads over the years.
As for the voters, be cautious with that argument. Lebron has four MVPs, while Kobe has one. Good luck persuading anyone that you get to pick and choose which votes count. Anyway, as I've said, advanced metrics are blowing up a lot of the defensive votes from years past.
Goldenwest wrote:
-like i said statistics only tell half the story. and your talking as if there's a big difference between Kobe and Lebron, there isn't. Lebron has an edge in FG% only because he's bigger and thus plays much closer to the rim. But double teams or facing a strong defender? Stats don't tell you about that. That's why watching the games are so important. I've been saying a million times, the stats are not normalized for opponent defensive intensity, teammates reactions or when in the game points are scored. These are big omission in the stats. People like you really pay way too much attention to stats and what your doing is misusing the data to support an opinion without giving much credence to the actual play. Many times people that haven't played much do this. Maybe You should get your own head out of the stat sheet.
Right, the stats are a conspiracy. We're all misusing them because they show that Kobe wasn't as great as you think he was. Sure thing.
I love Kobe, but his defense (much like Michael Jordan's) was VASTLY overrated and many of his All Defense Awards were a reach at best.
He certainly had the ability to play great defense (Game 7 vs. Boston being a prime example) but this is a guy who essentially developed the term "roamer" as he played a 1 man floating zone for 10 years in his 24 phase with rare exceptions. _________________ Love, Laker Lanny
Well, here's one area where I agree with Batguano. Every advanced statistic is subjective. Basic stats like FGM are not subjective.
But every single advanced statistic is subjective because advanced stats try to measure things that cannot be measured. Things like efficiency. Or defense.
So that means someone has to decide which variables, constants, multipliers, and so on should or should not be included in the advanced statistic. And then, when they create their initial formula, they have to run it and eyeball it and see if the results make sense to THEM, and if not, tweak the formula until the result set mirrors their own personal opinions about efficiency or defense.
That's been my whole point to him. If he's making the argument that things like "efficiency" are subjective, then why is he so hell bent on trying to argue who is "best". It's no different than Hollinger trying to argue who is most efficient. Doesn't matter whether you use an advanced statistic or observations or maybe neither or you just pick your favorite player and put him there, it's all ... subjective.
So while the effort to quote posts is admirable, it's fruitless. It doesn't mean anything. It doesn't MAKE a player the best anymore than Hollinger's PER makes the player with the higher score necessarily more efficient.
That's not actually true, though... which is my point. RAPM is just based on a simple premise: if a player is good, then his team will outscore the other team more often than when he's not there. The only tricky part is what to include as the prior, and for RAPM, using up to the last three years' past history (or just a "rookies are bad" in general) as the prior information.
The same thing is nearly true of BPM. They just took a bunch of different box score stats, figured out which ones could predict RAPM the best, and used those ones. No one hand chose any variables, constants. The only thing hand chosen is the multiplier for team record (which I've been critical of).
Not every stat is hand crafted like WS/48 and PER (and you'll note I call them bad stats... especially PER). That doesn't mean RAPM or BPM are flawless, not by a long shot. But your specific criticism doesn't actually apply to them.
Yes it does. So, RAPM essentially tries to determine "value".
That means someone has to decide which stats are to be included in the RAPM formula.
Which means, someone is deciding which stats are to be included in determining player value AND that that combination of stats does in fact determine value in some way.
Which stats get included in calculating RAPM, and to what degree, is entirely subjective. Just because you like the stat, or even if the stat is highly accurate, doesn't mean it isn't subjective.
Even eFG% and TS% are subjective. Someone had to decide for eFG% that FT% is not to be included and for TS%, that it should.
-Kobe was always a lock down defender taking on the most prolific scorer in crunch time. He was never an average defender pre-2012. So you know more than the voters? And anyway 'average to mediocre'?
Bull. Kobe was not a lock down defender, and he often was not matched up with the other team's star scorer. We've been through this in many other threads over the years.
As for the voters, be cautious with that argument. Lebron has four MVPs, while Kobe has one. Good luck persuading anyone that you get to pick and choose which votes count. Anyway, as I've said, advanced metrics are blowing up a lot of the defensive votes from years past.
Goldenwest wrote:
-like i said statistics only tell half the story. and your talking as if there's a big difference between Kobe and Lebron, there isn't. Lebron has an edge in FG% only because he's bigger and thus plays much closer to the rim. But double teams or facing a strong defender? Stats don't tell you about that. That's why watching the games are so important. I've been saying a million times, the stats are not normalized for opponent defensive intensity, teammates reactions or when in the game points are scored. These are big omission in the stats. People like you really pay way too much attention to stats and what your doing is misusing the data to support an opinion without giving much credence to the actual play. Many times people that haven't played much do this. Maybe You should get your own head out of the stat sheet.
Right, the stats are a conspiracy. We're all misusing them because they show that Kobe wasn't as great as you think he was. Sure thing.
Easy.
I value the ones voted on by head coaches over the ones voted on by the media.
I love Kobe, but his defense (much like Michael Jordan's) was VASTLY overrated and many of his All Defense Awards were a reach at best.
He certainly had the ability to play great defense (Game 7 vs. Boston being a prime example) but this is a guy who essentially developed the term "roamer" as he played a 1 man floating zone for 10 years in his 24 phase with rare exceptions.
I think what people are conflating, is what Kobe was capable of doing versus what he actually did.
I do believe Kobe was entirely capable of being a lockdown, Artest-level defender. But he didn't do that all that often because he saved much of that energy for the offensive side of the ball.
Believing, and arguably being right, that the net outcome would be more positive by doing so. Meaning, maybe if he did lockdown d he could prevent 8 points per game scoring 23 (net of +15 PPG) versus saving himself for offense so only preventing 4 points per game but scoring 25 (net of +21 PPG). Just an example.
-Kobe was always a lock down defender taking on the most prolific scorer in crunch time. He was never an average defender pre-2012. So you know more than the voters? And anyway 'average to mediocre'?
Bull. Kobe was not a lock down defender, and he often was not matched up with the other team's star scorer. We've been through this in many other threads over the years.
As for the voters, be cautious with that argument. Lebron has four MVPs, while Kobe has one. Good luck persuading anyone that you get to pick and choose which votes count. Anyway, as I've said, advanced metrics are blowing up a lot of the defensive votes from years past.
Goldenwest wrote:
-like i said statistics only tell half the story. and your talking as if there's a big difference between Kobe and Lebron, there isn't. Lebron has an edge in FG% only because he's bigger and thus plays much closer to the rim. But double teams or facing a strong defender? Stats don't tell you about that. That's why watching the games are so important. I've been saying a million times, the stats are not normalized for opponent defensive intensity, teammates reactions or when in the game points are scored. These are big omission in the stats. People like you really pay way too much attention to stats and what your doing is misusing the data to support an opinion without giving much credence to the actual play. Many times people that haven't played much do this. Maybe You should get your own head out of the stat sheet.
Right, the stats are a conspiracy. We're all misusing them because they show that Kobe wasn't as great as you think he was. Sure thing.
Easy.
I value the ones voted on by head coaches over the ones voted on by the media.
Does that mean you agree with the idea that Kobe should never have won a DPOY award?
-Kobe was always a lock down defender taking on the most prolific scorer in crunch time. He was never an average defender pre-2012. So you know more than the voters? And anyway 'average to mediocre'?
Bull. Kobe was not a lock down defender, and he often was not matched up with the other team's star scorer. We've been through this in many other threads over the years.
As for the voters, be cautious with that argument. Lebron has four MVPs, while Kobe has one. Good luck persuading anyone that you get to pick and choose which votes count. Anyway, as I've said, advanced metrics are blowing up a lot of the defensive votes from years past.
Goldenwest wrote:
-like i said statistics only tell half the story. and your talking as if there's a big difference between Kobe and Lebron, there isn't. Lebron has an edge in FG% only because he's bigger and thus plays much closer to the rim. But double teams or facing a strong defender? Stats don't tell you about that. That's why watching the games are so important. I've been saying a million times, the stats are not normalized for opponent defensive intensity, teammates reactions or when in the game points are scored. These are big omission in the stats. People like you really pay way too much attention to stats and what your doing is misusing the data to support an opinion without giving much credence to the actual play. Many times people that haven't played much do this. Maybe You should get your own head out of the stat sheet.
Right, the stats are a conspiracy. We're all misusing them because they show that Kobe wasn't as great as you think he was. Sure thing.
Easy.
I value the ones voted on by head coaches over the ones voted on by the media.
Does that mean you agree with the idea that Kobe should never have won a DPOY award?
That one is voted on by the media as well.
But I don't know if there's any year where I would've given Kobe DPOY. But that's cool. He has more than enough All-Defense selections to make up for it.
I love Kobe, but his defense (much like Michael Jordan's) was VASTLY overrated and many of his All Defense Awards were a reach at best.
He certainly had the ability to play great defense (Game 7 vs. Boston being a prime example) but this is a guy who essentially developed the term "roamer" as he played a 1 man floating zone for 10 years in his 24 phase with rare exceptions.
I think what people are conflating, is what Kobe was capable of doing versus what he actually did.
I do believe Kobe was entirely capable of being a lockdown, Artest-level defender. But he didn't do that all that often because he saved much of that energy for the offensive side of the ball.
Believing, and arguably being right, that the net outcome would be more positive by doing so. Meaning, maybe if he did lockdown d he could prevent 8 points per game scoring 23 (net of +15 PPG) versus saving himself for offense so only preventing 4 points per game but scoring 25 (net of +21 PPG). Just an example.
I don't think his defense was ever at the level of prime Artest, but yes, he was capable of playing very good defense when he wanted to. Yes, he chose not to. Yes, maybe it was a good decision on a net basis.
However, you can't evaluate his career on the basis of what he could have done if he had made different choices.
By the way, Lanny is 100% right about MJ's defense. It was overrated, too, and for the same reason. MJ let Pippen do most of the heavy lifting on defense. Toward the end of his time in Chicago, MJ was actually the weak link in the Bulls' defense (bearing in mind that the Bulls were a very good defensive team).
Yes it does. So, RAPM essentially tries to determine "value".
That means someone has to decide which stats are to be included in the RAPM formula.
Which means, someone is deciding which stats are to be included in determining player value AND that that combination of stats does in fact determine value in some way.
Which stats get included in calculating RAPM, and to what degree, is entirely subjective. Just because you like the stat, or even if the stat is highly accurate, doesn't mean it isn't subjective.
Even eFG% and TS% are subjective. Someone had to decide for eFG% that FT% is not to be included and for TS%, that it should.
I don't think you understand how RAPM works, but that's beside the point. When you say that even TS% is subjective, you're going off the rails. Isn't FG% itself subjective? I mean, someone had to decide to divide FGM by FGA. This becomes an exercise in low level epistemology -- how do we ever know anything? All of reality is subjective because it is colored by our perceptions.
The old school box score composite stats like PER had an underlying methodology. The real limitation on those stats is not the methodology, but instead the inherent limitations of box score stats. The point of PER and the like is to reduce a large number of box score stats into a single number. That has value, though no one (except maybe Hollinger at times) ever held those stats out as definitive. They were just convenient and a reasonable distillation of box score stats. That's why people still use them for some purposes.
Any serious analysis of statistics looks at a wide range of measures. Sure, the analysis still has some element of subjectivity, but that's true of just about everything in the real world, if you want to look at it that way.
Yes it does. So, RAPM essentially tries to determine "value".
That means someone has to decide which stats are to be included in the RAPM formula.
Which means, someone is deciding which stats are to be included in determining player value AND that that combination of stats does in fact determine value in some way.
Which stats get included in calculating RAPM, and to what degree, is entirely subjective. Just because you like the stat, or even if the stat is highly accurate, doesn't mean it isn't subjective.
Even eFG% and TS% are subjective. Someone had to decide for eFG% that FT% is not to be included and for TS%, that it should.
I don't think you understand how RAPM works, but that's beside the point. When you say that even TS% is subjective, you're going off the rails. Isn't FG% itself subjective? I mean, someone had to decide to divide FGM by FGA. This becomes an exercise in low level epistemology -- how do we ever know anything? All of reality is subjective because it is colored by our perceptions.
The old school box score composite stats like PER had an underlying methodology. The real limitation on those stats is not the methodology, but instead the inherent limitations of box score stats. The point of PER and the like is to reduce a large number of box score stats into a single number. That has value, though no one (except maybe Hollinger at times) ever held those stats out as definitive. They were just convenient and a reasonable distillation of box score stats. That's why people still use them for some purposes.
Any serious analysis of statistics looks at a wide range of measures. Sure, the analysis still has some element of subjectivity, but that's true of just about everything in the real world, if you want to look at it that way.
This is what people are learning about analytics in general.
The methodology (such as the algorithms) are not as important as the data that is fed into them. The problem with PER isn't a lack of logic or subjectivity, but simply that the data being fed into it (box score stats) isn't great. And no matter how good your methodology, you can only do so much with weak data.
Today, much better and more refined data is captured so the result is better.
There are always going to be who dismiss analytics, particularly when (a) they don't like the results or (B) they don't like being told that anything beyond their personal opinion and observations matter.
From marketers to people who run supply chains to sports coaches, people have bristled against analytics because they don't like having their opinion vetted or challenged.
It also generally correlates with quality of teammates, which could be interpreted a couple different ways. As for your point, I don't know whether the stats include the playoffs. However, I noticed that he didn't have a separate stat for the playoffs, so maybe they do.
That's a good point, though the 2011 cast wasn't any worse than the 2010 cast in the regular season (57 wins in both cases). But that can be explained as an outlier. I find it unlikely, since RAPM (in theory) accounts for effects of teammate quality. In practice, it's flawed when doing so, but it does a good enough job to make me doubt that's the primary driver of this trend.
It does include the playoffs. I don't have "proof" anywhere but I distinctly remember reading that it does.
Yes it does. So, RAPM essentially tries to determine "value".
That means someone has to decide which stats are to be included in the RAPM formula.
Which means, someone is deciding which stats are to be included in determining player value AND that that combination of stats does in fact determine value in some way.
Which stats get included in calculating RAPM, and to what degree, is entirely subjective. Just because you like the stat, or even if the stat is highly accurate, doesn't mean it isn't subjective.
Even eFG% and TS% are subjective. Someone had to decide for eFG% that FT% is not to be included and for TS%, that it should.
This is moving the goalposts. Your original claim was that "someone has to decide which variables, constants, multipliers, and so on" and that's patently not true.
Yes, obviously in RAPM, someone had to choose the general setup of the problem. So you can't just use RAPM without understanding how it was built. But after someone chose the general optimization problem, there were no arbitrary variables, constants etc. that were added to it.
This is BTW why I don't like RPM. Because it does add arbitrary variables and constants.
The methodology (such as the algorithms) are not as important as the data that is fed into them. The problem with PER isn't a lack of logic or subjectivity, but simply that the data being fed into it (box score stats) isn't great. And no matter how good your methodology, you can only do so much with weak data.
Today, much better and more refined data is captured so the result is better.
Both can be true. The problem with PER can be as ringfinger and others claimed (who's picking these formulas and under what rationale?) AND that the inputs to the stat don't capture the full range of basketball productivity. You can definitely come up with a better box score summary of a player than PER, but it's also true that no matter what summary you come up with, it'll always be flawed due to the limited box score stats.
1) Every single All-time great in every sport has their accolades "padded" in some way or another by getting a couple just based on reputation and all-time greatness towards the end of their career. Why is it such a big deal with Kobe?
2) All of Kobe's All-Defense selections were in an era were head coaches did the voting. Not saying they're perfect but give me the opinions of NBA head coaches over media-selected accolades any day. These guys spend hours watching these players up close in game situations and in game film, game-planning for them.
3) It's difficult to measure All-Defense selections because all the players awarded them are not created equal. On one end of the spectrum you have guys that are roleplayers and who's SOLE job is to focus all their energy on defense and shutting down the best player (Bowen, Tony Allen, Avery Bradley) and you have all-stars/HOFs who also fall under this category (Rodman, Ben Wallace, Mutombo). On the other end of the spectrum you have superstars that carry their teams offensively (Jordan, LeBron, Kobe) but are also capable of playing elite defense. How can you compare someone like Bowen who's sole focus is to play defense vs someone like Kobe who has to carry the offense and also play elite defense? Imagine if guys like Kobe and Jordan got to JUST focus on defense all game without having to score 30+ and be their teams primary playmaker. When it comes down to it there is a very short list of players in the history of the game who can score 40+ on any given night and also play elite defense (and Kobe is one of them)
4) Kobe was excellent defensively in the playoffs (when it counts the most). Examples: his defense during the 2000 title run, especially Game 7 vs Blazers in the 2000 WCFs; his defense against Bibby in Game 4 in the 2002 WCFs; his defense in the 2008 Olympics; his defense against Westbrook in the 2010 playoffs. These were game/series altering defensive performances. LeBron is stuck at 6 All-Defense selections because he's equally being accused of "coasting" on defense in the regular season, because of the offensive load he has to carry. But that doesn't stop him from being one of the greatest defensive players of all time, because everyone knows he turns it on in the playoffs.
5) The team selections are G, G, F, F, C. Not position-specific. So you're doing a bit of a disservice just saying that there simply weren't better SG defenders in the league when he's going up against all guards.
I don't think the media would have voted any different than the head coaches and vice versa. The MVPs, ROY's, defensive teams, etc are generally pretty aligned with popular opinion. If the media voted on defensive teams Kobe would have made it every year. Few media members ever questioned Kobe's defense.
1) Every single All-time great in every sport has their accolades "padded" in some way or another by getting a couple just based on reputation and all-time greatness towards the end of their career. Why is it such a big deal with Kobe?
2) All of Kobe's All-Defense selections were in an era were head coaches did the voting. Not saying they're perfect but give me the opinions of NBA head coaches over media-selected accolades any day. These guys spend hours watching these players up close in game situations and in game film, game-planning for them.
3) It's difficult to measure All-Defense selections because all the players awarded them are not created equal. On one end of the spectrum you have guys that are roleplayers and who's SOLE job is to focus all their energy on defense and shutting down the best player (Bowen, Tony Allen, Avery Bradley) and you have all-stars/HOFs who also fall under this category (Rodman, Ben Wallace, Mutombo). On the other end of the spectrum you have superstars that carry their teams offensively (Jordan, LeBron, Kobe) but are also capable of playing elite defense. How can you compare someone like Bowen who's sole focus is to play defense vs someone like Kobe who has to carry the offense and also play elite defense? Imagine if guys like Kobe and Jordan got to JUST focus on defense all game without having to score 30+ and be their teams primary playmaker. When it comes down to it there is a very short list of players in the history of the game who can score 40+ on any given night and also play elite defense (and Kobe is one of them)
4) Kobe was excellent defensively in the playoffs (when it counts the most). Examples: his defense during the 2000 title run, especially Game 7 vs Blazers in the 2000 WCFs; his defense against Bibby in Game 4 in the 2002 WCFs; his defense in the 2008 Olympics; his defense against Westbrook in the 2010 playoffs. These were game/series altering defensive performances. LeBron is stuck at 6 All-Defense selections because he's equally being accused of "coasting" on defense in the regular season, because of the offensive load he has to carry. But that doesn't stop him from being one of the greatest defensive players of all time, because everyone knows he turns it on in the playoffs.
5) The team selections are G, G, F, F, C. Not position-specific. So you're doing a bit of a disservice just saying that there simply weren't better SG defenders in the league when he's going up against all guards.
I don't think the media would have voted any different than the head coaches and vice versa. The MVPs, ROY's, defensive teams, etc are generally pretty aligned with popular opinion. If the media voted on defensive teams Kobe would have made it every year. Few media members ever questioned Kobe's defense.
good stuff batg, again.
the only criticism of kobe's D is that he took games off. but this is a common thing said about everyone. it's not that true when you factor in all the responsibilities kobe had and the era he played in. MJ was not really different about taking plays off, etc. all primary scorers will HAVE to do this to a significant degree. Someone like rodman never took plays off, but he also wasn't relied on AT ALL for offense...so just defense and rebound. it's easy to do that, very easy job compared to trying to score. ben wallace, dwight, are all like this. no offense, so they focus on D almost exclusively.
that aside, kobe's defense was extremely high level, elite stuff. remember the all star game? when kobe keyed on lebron, lebron couldn't do jack, and had to be bailed out within a minute or two by the refs. kobe vs kyrie, remember that? kobe. so many examples in the playoffs too. and how about olympics? when he didn't need to think about scoring. he shut barbosa down one game when he was leading the olympics in scoring. there's a reason why all these players talk about kobe so highly. every aspect of the game he has mastered. i think wade called him like a black belt basketball master once.
Both can be true. The problem with PER can be as ringfinger and others claimed (who's picking these formulas and under what rationale?) AND that the inputs to the stat don't capture the full range of basketball productivity. You can definitely come up with a better box score summary of a player than PER, but it's also true that no matter what summary you come up with, it'll always be flawed due to the limited box score stats.
About 10 years ago, I picked up a copy of the issue of Basketball Prospectus in which Hollinger laid out his formula for PER and its rationale. I thought that the only aspect of the formula that was reasonably debatable was his treatment of assists. Over the years, I've seen people take shots at PER. Most of the time, this has been more of a reaction to Hollinger than to PER itself. I've looked at some of the composite box score stats that have been offered as alternatives to PER, and frankly none of them struck me as substantially different.
It does include the playoffs. I don't have "proof" anywhere but I distinctly remember reading that it does.
Then your explanation makes sense. My alternative explanation about teammates is not based on the idea that his teammates made him look better, but rather that he may have put more emphasis on his defense once he had better teammates. Around 2006, he was channeling his effort into scoring.
Both can be true. The problem with PER can be as ringfinger and others claimed (who's picking these formulas and under what rationale?) AND that the inputs to the stat don't capture the full range of basketball productivity. You can definitely come up with a better box score summary of a player than PER, but it's also true that no matter what summary you come up with, it'll always be flawed due to the limited box score stats.
About 10 years ago, I picked up a copy of the issue of Basketball Prospectus in which Hollinger laid out his formula for PER and its rationale. I thought that the only aspect of the formula that was reasonably debatable was his treatment of assists. Over the years, I've seen people take shots at PER. Most of the time, this has been more of a reaction to Hollinger than to PER itself. I've looked at some of the composite box score stats that have been offered as alternatives to PER, and frankly none of them struck me as substantially different.
The problem with ostensibly "reasonable" rationale is that just because something seems plausible doesn't make it actually correct.
1) The stats it chooses to highlight aren't necessarily right. When you work with numbers, you find that your intuition often ends up being wrong. Hollinger might think that his stats might make sense. But how do we know that, say, 3 point attempt rate (3Ar) isn't also a key factor? (OBPM thinks it is?) And if it is, is it linear or is it maybe logarithmic? Do I scale it by USG%? Maybe by 3FG%?
2) The coefficients it chooses aren't necessarily right. Let's say that Hollinger did use the best set of stats (which he probably didn't). How do you know what the coefficients should be? How do they relate? He's placing his own belief in which stats have value here... but who is John Hollinger to dictate what is right?
That's why I stress BPM is a far better all-in-one box score stat, even as guys like Westbrook have broken the stat recently. BPM bypasses the above two problems by making a simple assumption: "Matching RAPM's ratings closely is our target." Based on that, authors can choose any set of variables, let the algorithm (i.e. regression) find the best coefficients, and compute residuals to find out whether it's a good match. That's how you end up with something DRB% * AST% interaction playing a big role in DBPM -- something I've been critical of in the past (vis-a-vis WB, Harden, Jokic, Randle) but also something that empirically has helped distinguish good defenders (LeBron, KG, Marc Gasol). They simply found that including that term made a better fit to RAPM, so they kept it in -- as opposed to relying on the unilteral judgment of one statistician (and not a particularly impressive one)
You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum
