Talk Me Down: How is Deng not MVP of the Bulls?

[The Evidence]

http://godismyjudgeok.com/DStats/aspm-a ... 2012-aspm/

According to this, Rose beats Deng in ASPM (both oASPM & dASPM ) and VORP pretty handily.

[db42]

For dASPM, the more negative, the better; so no, Rose doesn't beat Deng there.

Also, those stats say Carlos Boozer has a greater defensive impact than Deng. Not really buying that.

[SideshowBob]

Someone correct me if I'm wrong, but isn't SPM purely box score based? Why would you go for that over APM, particularly for defensive comparisons?

Take this as my ignorance of the SPM stat rather than an attack on it

[Doctor MJ]

Someone correct me if I'm wrong, but isn't SPM purely box score based? Why would you go for that over APM, particularly for defensive comparisons?

Take this as my ignorance of the SPM stat rather than an attack on it

I don't know what all he puts into ASPM, but yes, SPM is essentially just a PER-style metrics where the weights are determined by general correlation to the scoreboard. Given that the whole reason this thread exists is that Deng isn't impressive by the box score, the fact that he isn't impressive by a variant of SPM probably doesn't shed new light on the situation.

[SideshowBob]

Someone correct me if I'm wrong, but isn't SPM purely box score based? Why would you go for that over APM, particularly for defensive comparisons?

Take this as my ignorance of the SPM stat rather than an attack on it

I don't know what all he puts into ASPM, but yes, SPM is essentially just a PER-style metrics where the weights are determined by general correlation to the scoreboard. Given that the whole reason this thread exists is that Deng isn't impressive by the box score, the fact that he isn't impressive by a variant of SPM probably doesn't shed new light on the situation.

Thanks Doc.

Not to take away from Daniel's work, but I figured that ASPM wouldn't properly capture what Deng was doing just as any other boxscore limited stat wouldn't

[Sebastian]

Doc, you better have Deng as a top 2 DPOY come year's end with all this chatter!

(Not going to add anything new to the discussion, but I must say I like the tenor of the thread - more an intellectual pursuit of knowledge than a cro-magnon back-and-forth ****, which we see far too much of, especially in threads where such controversial opinions are argued).

[rrravenred]

I love this thead.

FWIW, Deng is one of the prototypes of what I like to call "fulcrum" players, who give all the other players in the team leverage to produce at a high level. A certain exercise of their skills (or a combination of skills) frees up other players to perform, either by spacing, help defense, ball-handling or even just the ability to position themselves correctly on court to allow other players to gamble, or hold pat, opening up an avenue for a player to slash to the basket or provide the threat of a jumper to allow a post player to work unmolested. As with Deng, or Battier (or even players like Nene or AK), their own production is less important to how they allow others to produce for team outcomes.

(an interesting point... Joe Johnson and Derek Fisher have almost identical 10-year RAPMs, on both sides, despite widely varying statistical production)

So does that mean that Deng is a BETTER player than Rose? Hardly. But it does mean that for the Bulls, the roles and positions that he fills on the team are more crucial to team success.

[ronnymac2]

I love this thead.

FWIW, Deng is one of the prototypes of what I like to call "fulcrum" players, who give all the other players in the team leverage to produce at a high level. A certain exercise of their skills (or a combination of skills) frees up other players to perform, either by spacing, help defense, ball-handling or even just the ability to position themselves correctly on court to allow other players to gamble, or hold pat, opening up an avenue for a player to slash to the basket or provide the threat of a jumper to allow a post player to work unmolested. As with Deng, or Battier (or even players like Nene or AK), their own production is less important to how they allow others to produce for team outcomes.

(an interesting point... Joe Johnson and Derek Fisher have almost identical 10-year RAPMs, on both sides, despite widely varying statistical production)

So does that mean that Deng is a BETTER player than Rose? Hardly. But it does mean that for the Bulls, the roles and positions that he fills on the team are more crucial to team success.

I see it in an almost the opposite way. Well, kind of.

I definitely see Deng being in the fulcrum role, where his jack-of-a-bunch-of-trades (he's not a jack-of-all-trades since he doesn't handle the ball a lot, even in a secondary role) style allows others on the team to concentrate on their specialities. That's all well and good.

But without Rose taking on so much offensive responsibility (lead volume guy, primary ball-handler, primary creator, primary floor general, primary attacker), Deng's contributions lose some of their value FOR CHICAGO.

Both players obviously have value. I'm not saying Rose's presence creates all of Deng's impact, because Deng clearly is a really good player who would impact games even if he was on a different team.

But if you look at the Bulls and ask "Who is the base on which they are built?", I'm not sure how it isn't Rose. Even if you analyze the team to see which positions are scarce in talent (So for example, Kobe is really valuable right now because most of L.A.'s high-end talent is concentrated in full court positions and roles while being weak on the perimeter), Rose still has great value since nobody else can move the ball via the live dribble consistently.

That's where Derrick's value lies.



I feel as though a lot of stats measure efficiency with which a player performs his role. Those stats are cool, but...what stat measures importance of a certain role itself? Not just how well one performs in it, but the importance of the role itself. I don't think there are stats for that. At least, not yet.


As an aside: If Chicago's only improvement offensively is through offensive rebounding, they don't have a shot against Miami. Personally, when the playoffs come around, I think Chicago will be better at simply executing offense compared to last year, so they'll have a chance. But if the strategy is just "Rose plus O-boards" again, it won't work.

[tsherkin]

Well, remember, their major addition isn't playing. Rip's been riding pine for a while and they've been counting on him to add another ball-handler (which he did reasonably well while healthy) and another attention-demanding scoring threat. Even though he wasn't hitting 3s or drawing fouls, he was moving the pieces of the defense around and drawing attention, then sticking mid-range jumpers all over the floor.

If and when he comes back, well, then that does change the dynamic of Chicago's offense a lot.

[colts18]

Tsherkin, how do you rank the Thunders top 4 players? How close are Collison and Harden to the stars?

[tsherkin]

Tsherkin, how do you rank the Thunders top 4 players? How close are Collison and Harden to the stars?

Collison is one of my favorite roleplayers ever. He is, though, a roleplayer. He could be better on the glass, but he is pure hustle and plays some fantastic defense. Typically a very good offensive rebounder, though.

Harden should be a starter by this point, he's extremely talented. I can see why Brooks wants him bolstering the second unit, because they don't have a lot else there, but he's got the skills of a potential All-Star.

Durant is ahead of them all, and I rank Harden ahead of Westbrook because Russell is not the brightest of players, though he is very talented, is legitimately a good player when he's not being an idiot and has come a long way since entering the league.

But that's more of another thread.

[rrravenred]

I see it in an almost the opposite way. Well, kind of.

What was that old quote that when you put two economists in a room you'll have at least three different positions within a minute?

I definitely see Deng being in the fulcrum role, where his jack-of-a-bunch-of-trades (he's not a jack-of-all-trades since he doesn't handle the ball a lot, even in a secondary role) style allows others on the team to concentrate on their specialities. That's all well and good.

But without Rose taking on so much offensive responsibility (lead volume guy, primary ball-handler, primary creator, primary floor general, primary attacker), Deng's contributions lose some of their value FOR CHICAGO.

Why, then, has Rose's absence not been problematic for the operations of the Bulls? That's (I suppose) the crux of the thread. Why are Rose's contributions relatively replaceable? Why is his volume able to be distributed across the roster on the floor, how do shots get created when he's off, how does the ball move, how does the defence come under pressure?

Please note, I'm not attacking or downplaying Rose as a player, merely asking why his (excellent) contributions as you've listed them don't appear to be missed quite as you'd expect.

It's also worth noting that what you've listed are all offensive attributes, which is not (IMO) the side that has defined the Bulls during the Thibs era.

Both players obviously have value. I'm not saying Rose's presence creates all of Deng's impact, because Deng clearly is a really good player who would impact games even if he was on a different team.

But if you look at the Bulls and ask "Who is the base on which they are built?", I'm not sure how it isn't Rose. Even if you analyze the team to see which positions are scarce in talent (So for example, Kobe is really valuable right now because most of L.A.'s high-end talent is concentrated in full court positions and roles while being weak on the perimeter), Rose still has great value since nobody else can move the ball via the live dribble consistently.

That's where Derrick's value lies.

FWIW, I think the Bulls are built on an ensemble base, rather than being slavishly dependent to any one player or player role on either side of the ball. A champion team will always beat a champion's team.

I feel as though a lot of stats measure efficiency with which a player performs his role. Those stats are cool, but...what stat measures importance of a certain role itself? Not just how well one performs in it, but the importance of the role itself. I don't think there are stats for that. At least, not yet.

Interesting point. With more detailed analysis of something like Synergy stats you could maybe get somewhere, but all the off-ball stuff (which is crucial) probably wouldn't get recorded anyway. But then you get into what roles are defined in different systems, where it gets REALLY messy. The PG role in Thibs', Phil's and Sloan's systems are so radically different (which, in a cascading effect, changes the roles of all the other players on the floor) that if you start trying to do cross-team comparisons it becomes a nightmare of trying to weight the different aspects of each role...

As an aside: If Chicago's only improvement offensively is through offensive rebounding, they don't have a shot against Miami. Personally, when the playoffs come around, I think Chicago will be better at simply executing offense compared to last year, so they'll have a chance. But if the strategy is just "Rose plus O-boards" again, it won't work.

Mmh... it'll partially depend on whether or not the Bulls offensive system can develop some flexibility to respond to the Heat throwing different defensive line-ups at them in pressure situations. But yeah, increased O-Rebounding won't cut it as a strategy.

[tsherkin]

Why, then, has Rose's absence not been problematic for the operations of the Bulls? That's (I suppose) the crux of the thread. Why are Rose's contributions relatively replaceable? Why is his volume able to be distributed across the roster on the floor, how do shots get created when he's off, how does the ball move, how does the defence come under pressure?

Quality of competition, or lack thereof, in his absence has been favorable to the Bulls.

A 10-game sample isn't exactly wonderful for evaluating everything that needs be evaluated in this context. He missed only one game last season, so that's 11 total games without him as a sample, and a weak one at that.

We actually don't have any idea how the Bulls would hold up without Rose over an extended period of time spent facing competent competition. Evaluating how well they do against the dregs of the league with a team that can defend and hit the boards isn't really a good thing to do. You're running into basically the 94 Bulls in that sense, a team that's extremely talented but lacking the drive and direction provided by its primary initiator. Talented enough to get by against weaker competition but ultimately lacking in critical punch against top teams. Obviously on a lesser scale than those specific Bulls, but used as an analogy, I feel it an effective comparison.

[mysticbb]

Someone correct me if I'm wrong, but isn't SPM purely box score based? Why would you go for that over APM, particularly for defensive comparisons?

Take this as my ignorance of the SPM stat rather than an attack on it

I don't know what all he puts into ASPM, but yes, SPM is essentially just a PER-style metrics where the weights are determined by general correlation to the scoreboard. Given that the whole reason this thread exists is that Deng isn't impressive by the box score, the fact that he isn't impressive by a variant of SPM probably doesn't shed new light on the situation.

unadjusted SPM = a*MPG + b*TRB% + c*BLK% + d*STL% + e*USG%*[TS%*2*(1-TO%) – f*TO% – g + h*AST% + i*USG%]

The coefficients a, b, c, d, e, f, g, h and i are found via baysian regression of the independent variables MPG, TRB%. BLK%, STL%, USG%, TS%, TO%, AST% on a 8 year unweighted RAPM.

After that an team and opponents adjustment is made.

And yes, Deng doesn't produce the boxscore stats in order to look good by any metric based on boxscore stats. Deng's value is basically seen in the results with him on the court and especially on defense. He might be the player with the least amount of defensive mistakes, the best in terms of overall team defense and help. The Bulls are playing so much better defensively with him on the court and all that while he plays so many minutes, it is amazing. The Bulls with Deng and without Rose on the court are playing better than the Bulls with Rose and without Deng. Overall Deng is more important to the overall team success and it can be said that Deng is the MVP of the Bulls in this season.

[Wharton Alum 08]

Someone correct me if I'm wrong, but isn't SPM purely box score based? Why would you go for that over APM, particularly for defensive comparisons?

Take this as my ignorance of the SPM stat rather than an attack on it

I don't know what all he puts into ASPM, but yes, SPM is essentially just a PER-style metrics where the weights are determined by general correlation to the scoreboard. Given that the whole reason this thread exists is that Deng isn't impressive by the box score, the fact that he isn't impressive by a variant of SPM probably doesn't shed new light on the situation.

unadjusted SPM = a*MPG + b*TRB% + c*BLK% + d*STL% + e*USG%*[TS%*2*(1-TO%) – f*TO% – g + h*AST% + i*USG%]

The coefficients a, b, c, d, e, f, g, h and i are found via baysian regression of the independent variables MPG, TRB%. BLK%, STL%, USG%, TS%, TO%, AST% on a 8 year unweighted RAPM.

After that an team and opponents adjustment is made.

And yes, Deng doesn't produce the boxscore stats in order to look good by any metric based on boxscore stats. Deng's value is basically seen in the results with him on the court and especially on defense. He might be the player with the least amount of defensive mistakes, the best in terms of overall team defense and help. The Bulls are playing so much better defensively with him on the court and all that while he plays so many minutes, it is amazing. The Bulls with Deng and without Rose on the court are playing better than the Bulls with Rose and without Deng. Overall Deng is more important to the overall team success and it can be said that Deng is the MVP of the Bulls in this season.

I'm pretty sure Luol Deng will disagree with this statement.

[Sebastian]

Someone correct me if I'm wrong, but isn't SPM purely box score based? Why would you go for that over APM, particularly for defensive comparisons?

Take this as my ignorance of the SPM stat rather than an attack on it

I don't know what all he puts into ASPM, but yes, SPM is essentially just a PER-style metrics where the weights are determined by general correlation to the scoreboard. Given that the whole reason this thread exists is that Deng isn't impressive by the box score, the fact that he isn't impressive by a variant of SPM probably doesn't shed new light on the situation.

unadjusted SPM = a*MPG + b*TRB% + c*BLK% + d*STL% + e*USG%*[TS%*2*(1-TO%) – f*TO% – g + h*AST% + i*USG%]

The coefficients a, b, c, d, e, f, g, h and i are found via baysian regression of the independent variables MPG, TRB%. BLK%, STL%, USG%, TS%, TO%, AST% on a 8 year unweighted RAPM.

After that an team and opponents adjustment is made.

And yes, Deng doesn't produce the boxscore stats in order to look good by any metric based on boxscore stats. Deng's value is basically seen in the results with him on the court and especially on defense. He might be the player with the least amount of defensive mistakes, the best in terms of overall team defense and help. The Bulls are playing so much better defensively with him on the court and all that while he plays so many minutes, it is amazing. The Bulls with Deng and without Rose on the court are playing better than the Bulls with Rose and without Deng. Overall Deng is more important to the overall team success and it can be said that Deng is the MVP of the Bulls in this season.

So if he is the anchor of arguably the best team in the NBA, does that make him a legitimate MVP candidate in your eyes?

[mysticbb]

So if he is the anchor of arguably the best team in the NBA, does that make him a legitimate MVP candidate in your eyes?

In terms of overall per game impact? Yes. But he also missed 7 games so far AND LeBron James is playing on a really high level. That makes James my choice of the MVP, but Deng should get considerations, if he can keep his playing level while not missing more games.

[drza]

Why, then, has Rose's absence not been problematic for the operations of the Bulls? That's (I suppose) the crux of the thread. Why are Rose's contributions relatively replaceable? Why is his volume able to be distributed across the roster on the floor, how do shots get created when he's off, how does the ball move, how does the defence come under pressure?

Quality of competition, or lack thereof, in his absence has been favorable to the Bulls.

A 10-game sample isn't exactly wonderful for evaluating everything that needs be evaluated in this context. He missed only one game last season, so that's 11 total games without him as a sample, and a weak one at that.

We actually don't have any idea how the Bulls would hold up without Rose over an extended period of time spent facing competent competition. Evaluating how well they do against the dregs of the league with a team that can defend and hit the boards isn't really a good thing to do. You're running into basically the 94 Bulls in that sense, a team that's extremely talented but lacking the drive and direction provided by its primary initiator. Talented enough to get by against weaker competition but ultimately lacking in critical punch against top teams. Obviously on a lesser scale than those specific Bulls, but used as an analogy, I feel it an effective comparison.

While true that an 11 game sample isn't enough, those that find value in the APM stats would point out that we're looking at a much larger sample than just 11 games. Every game there are on the order of 20 different 5-man line-ups that play for a team, which gives us several snapshots from every game of the player in question (here, Rose or Deng) with a bunch of other groups of 4 men.

Quick aside, I think this is one of the areas of APM that not everyone latches onto. It's not just about how a team plays with one player on the court vs when he's not on the court. It's about how each 4-man combination does with him as opposed to with someone else, which gives much more added and info for much more nuanced conclusions. Without going too abstract, and keeping it to Rose and Deng, it might look something like:

Rose/Rip/Deng/Boozer/Noah --> 500 minutes, +50
Rose/Rip/Brewer/Boozer/Noah --> 100 minutes, +5
Lucas/Rip/Deng/Boozer/Noah --> 100 minutes, +20
Rose/Rip/Korver/Boozer/Noah --> 100 minutes, +1
Watson/Rip/Deng/Boozer/Noah--> 100 minutes, +29
James/Rip/Deng/Boozer/Noah --> 100 minutes, +17
etc.

Obviously I was just choosing round numbers to make the point clear, but when you look at all of the different combos just exchanging a single player you can really start to isolate how much each individual player contributes. In other words, you're not just looking at how much one player lifts a certain group...or how much the absence of a certain player hurts a certain group...you're instead solving for exactly what one player's estimated contribution is, precisely, to the team's success and failure. It's like algebra...solving for multiple unknowns with multiple equations. I'm not sure that is really intuitive to folks that haven't really put in the time to understand the mechanisms of APM, vs. the much more intuitive on/off approach.

The problem with what I just wrote, of course, is that in any single year you rarely get a large enough sample size with all of the different groups in enough of a representative cross section of situations to solve very accurately. Bringing it back to the algebra, in a given season you generally end up with more unknowns than equations and have to estimate. The more estimates you have to make, the more potential error you're introducing to your conclusions. Which is why folks have started looking at either multiple years at a time to get that large enough sample size to minimize the # of estimates and/or trying math tricks like "Regularizing" to make the estimates you do have to make more accurate.

Which, again, brings us back to Deng and Rose. The fact that Deng has the consistently super-high APM going all the way back to the '04 study that captures his rookie year, in different multi-year studies, tells me that when you look at the vast number of 5-man teams and situations that Deng has played in, he specifically can be pointed to as very definite positive value added. This isn't a case where he's just a "lucky" role player that always happens to be out there on a team where he fits best...this is Deng's 8th season, he's played for 5 different coaches, on teams ranging from 33 wins up to 62 wins. He played four full seasons before Derrick Rose ever stepped foot on an NBA court, with point guards like Hinrich or Duhon that weren't exactly offensive juggernauts. And he still CONSISTENLY has a very measurable, very positive impact on every team while playing major minutes.

Now, tsherkin had a good point earlier about the potential evolution of Rose's game...if Rose's game has changed significantly this season compared to last, APM won't have had enough data yet to establish his current level. APM is best as a multi-year review, not an instantaneous measure. And Rose obviously made major changes last season compared to his first couple, and he's very young, so he could still be evolving.

But that said, Deng has a very established track record of very positive (and measurable) impact on a team with or without Rose. And though 11 games is too small of a sample size to draw a conclusion on Rose's relative importance, the fact that the results of those 11 games matches very well with what we'd have expected from a Deng-led talented team based on the thousands of minutes from 7 years worth of APM results...I dunno. I'll say the jury is still out on Rose's impact, but I'm pretty comfortable with Deng's.

[Doctor MJ]

The fact that Deng has the consistently super-high APM going all the way back to the '04 study that captures his rookie year, in different multi-year studies,

Please list more on that. The data I have previous gathered makes me hesitant to go too far with that.

Ilardi's 6-year from '03-'09 ranked him at 48th.
Engelmann's 5-year study he had on his side mid-season last year had Deng at 27th.

I do realize that more recent studies see climbing significantly even with many year studies, but the truth is that until recently, even the multi-year studies I've seen only saw Deng as a very good player, not what I'd call super-high.

Of course, more recent events SHOULD be able to sway rankings, but it seems really clear at the very least that we should not be looking like Deng as a guy who has been a secret superstar forever. In fact, he wasn't even a guy championed by +/- stat folks really until recently.

[Doctor MJ]

I'm pretty sure Luol Deng will disagree with this statement.

The funny thing is: There's no reason to expect that Deng (or anyone else for that matter) really has that accurate of a gauge of his own impact relative to someone else playing an entirely different role.