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

[alucryts]

What is the percent error based on that standard deviation?

percent error = (standard deviation / standard value) * 100

If this number is as massive as I think it is going to be, your error is as big or bigger than the value itself. If this is the case, it is statistically favorable that anyone near zero plus minus on the team during those stretches can be the DPOY according to your logic. The mean value no longer means anything when the standard deviation is so large relatively.

[alucryts]

It is very easy to remove Deng from the defense on offense. If he was truly the cause of the great defense, offensive teams would remove him from the play and go to work on our defense. You can easily remove one perimeter player's impact from a defense. Teams don't remove perimeter players from defense because they are not the cause of great defense; they only help it.

Are you sure this is true? Hasn't your stance been that Deng is the Bull most likely to be covering the LeBron's, the Durant's, the Pierce's, the Grangers, etc. on defense? If that's the case, the only way for an offense to remove Deng is to essentially remove one of their best (if not the best) offensive option from their game plan? And if they did, wouldn't their overall offensive performance likely suffer in that instance?

In other words, haven't you set up a case where either Deng could make a really large defensive impact on his man, or he could make a similar (maybe even bigger) defensive impact by completely removing his opponent from the offense's game plan? Or have I misinterpreted what you mean here?

Not entirely the case. It isn't something that has to happen every time down the floor, but if the argument for Deng is based around help defense, then making him a useless help defender is very easy. I should have stated it is easy to remove his impact as a help defender than entirely from the defense because your point is valid. You cannot remove him from the defense entirely, but you can abolish his help defense. With this line of thinking and considering the Bulls defensive scheme, if it is both easy to remove his help defense and the Bulls do not rely on 1 on 1 perimeter defense almost ever, where is his DPOY impact that is caused by his incredible help defense? Either his help defense is not what causes the Bulls defense to tick, or teams simply don't respect him as the Bull's defensive centerpiece.

[Dr Positivity]

The reason I have a hard time accepting Deng as a DPOY type defender is the type of perimeter D I usually favor in the post handcheck league is rotations as fast and hard as possible. To me what makes the most sense pragmatically is that the perimeter defenders who will play the best help D are the ones who can cover the most ground and fastest. That's why I'd have a hard time accepting Deng is a better defender than Lebron, Wade, Iguodala. The actual reasons for why Deng would be better than someone like Lebron or Iguodala defensively doesn't check out to me. Raw intelligence? It's not like a LBJ and Iguodala don't know they're doing. I see their speed as more valuable. LBJ's physical talent just kills teams defensively. I think Deng is top 5-6 for perimeter D. Something like Lebron and Iguodala 1 2, then Wade, G Wallace, Deng and Tony Allen in some order

Deng vs someone like Dwight or Bogut or Chandler defensively just doesn't make that much sense to me. Maybe vs replacement you can make an argument. Though it's as hard to get an impact defensive big as an impact defensive perimeter player, so maybe not

[alucryts]

The reason I have a hard time accepting Deng as a DPOY type defender is the type of perimeter D I usually favor in the post handcheck league is rotations as fast and hard as possible. To me what makes the most sense pragmatically is that the perimeter defenders who will play the best help D are the ones who can cover the most ground and fastest. That's why I'd have a hard time accepting Deng is a better defender than Lebron, Wade, Iguodala. The actual reasons for why Deng would be better than someone like Lebron or Iguodala defensively doesn't check out to me. Raw intelligence? It's not like a LBJ and Iguodala don't know they're doing. I see their speed as more valuable. LBJ's physical talent just kills teams defensively. I think Deng is top 5-6 for perimeter D. Something like Lebron and Iguodala 1 2, then Wade, G Wallace, Deng and Tony Allen in some order

Deng vs someone like Dwight or Bogut or Chandler defensively just doesn't make that much sense to me. Maybe vs replacement you can make an argument. Though it's as hard to get an impact defensive big as an impact defensive perimeter player, so maybe not

Good post. I think that by the nature of the league, centers and help big men are always going to have more defensive impact than perimeter players. I do not think Deng is as good a defender if he does not have the ability to overplay the shot and the pass while lowering his attention on the drive because of the big men behind him.

[mysticbb]

If stats come to the conclusion that Deng is as defensively as valuable as Dwight Howard (and the only thing separating them is games missed), then you are interpreting the stats incorrectly; it is as simple as that. Deng is a great defensive player, but he is no DPOY.

Well, Deng is no DPOY, because you don't believe that he is? Abd that is your argument to say that I'm interpreting the stats incorrectly? Why do you think the stats based predictions are much better than yours? Because I interpret the stats incorrectly or you are just trying to ignore the facts?

It is very easy to remove Deng from the defense on offense. If he was truly the cause of the great defense, offensive teams would remove him from the play and go to work on our defense. You can easily remove one perimeter player's impact from a defense. Teams don't remove perimeter players from defense because they are not the cause of great defense; they only help it.

Maybe it is not as easy as you think to circumvent Deng's defensive influence? Usually the impact of a perimeter player on defense is indeed lower than by a big, but Noah is not the greatest individual defender anyway, that lowers his impact on that end. Additional to that you still think that Noah is the key here, but in the end you can replace Noah with basically any big and will see similar results, because, as you noted, Deng is one of the best to force offensive players into the help. If the perimeter player fails at that, he will lower his impact, Deng does not fail at that and can increase his impact with that. You are the one confusing cause and effect here. And that's why the results with Deng and without Noah are not only not worse defensively, but indeed better than without Deng and with Noah. The replacements for Deng are worse at that, and both, Korver and Brewer, aren't even that bad overall at the end. But it seems you are underestimating Deng's impact by giving out credit to the wrong player. If Noah would be the key here, it should not have much effect on the result when replacíng Deng by Korver or Brewer. If Deng is the cause here, it should have a great effect and replacing Noah should not make a big difference. And if we look at the lineup performances with Noah in comparison the performance with another player, it certainly looks the way as if Noah does not make a big difference afterall: http://stats-for-the-nba.appspot.com/pm/686.html

Can you imagine that you are placing value to the wrong thing here in your video analysis?

[mysticbb]

What is the percent error based on that standard deviation?

Given the fact that the Bulls DRtg is around 100, you should be able to calculate that by yourself. ;)

For the 82 games it is basically the same with 6.3%, and for the 7 games it is 5%.

[alucryts]

I was asking the standard deviation that leads to saying Deng provides +xx point on defense or the error associated with saying he is equal to Dwight. If plus minus only had a 5% error over 7 games, we would worship it as the god stat.

You keep implying that plus minus is fact when it is indeed not. Correlation does not imply caustion which means our correlations do not imply facts. How is replacing Deng with Korver equal to replacing Noah with Kurt Thomas or Asik?

Also, Rose and Brewer are also MVPs of our defense by your logic. They have job that Deng cannot do as well just as it is in reverse.

[mysticbb]

I was asking the standard deviation that leads to saying Deng provides +xx point on defense or the error associated with saying he is equal to Dwight.

+/- in that sense is just removing a fixed value for the average, which does not change the standard deviation. It is kind of silly to use that like you want it to be used. ;)

You keep implying that plus minus is fact when it is indeed not. Correlation does not imply caustion which means our correlations do not imply facts.

Raw +/- is a fact, those are real results completely unadjusted. Do you honestly want to imply that those are not facts?
And overall that all has not much to do with "correlation does not imply causation". We are not comparing two unrelated statistical variables here. We have a hypothesis, which can be refuted. We have context and we can test the ability to predict. And so far the predictions seem to confirm the results of that "correlation" while they are not in agreement with your argumentation.

Again, how do you explain that the value of Noah is not seen in the results? If he is so much more important than Deng, the effect should be seen with a different SF too. But it is not. Noah with someone different than Deng at SF creates a worse defense than Deng with someone else than Noah. Why is that, if Noah is the key?

Also, Rose and Brewer are also MVPs of our defense by your logic. They have job that Deng cannot do as well just as it is in reverse.

What? Seriously, my argumentation includes the results and the adjusted values. In which of those things are Rose or Brewer now at Deng's level in order to use my logic to come up with Rose and Brewer as MVP of the defense?

[lorak]

BTW, alucryts, last season you were using synergy stats to prove how good defensively Rose is.
So let's look at this year Bulls synergy stats:

ppp (rank among all players)

Asik 0,67 (17)
Deng 0,71 (32)
Gibson 0,71 (32)
Brewer 0,76 (74)
Rose 0,77 (85)
Noah 0,83 (167)
Boozer 0,95 (340)

Deng's result is even more impressive if we realize how tough defensive assignments he usually has.

[alucryts]

BTW, alucryts, last season you were using synergy stats to prove how good defensively Rose is.
So let's look at this year Bulls synergy stats:

ppp (rank among all players)

Asik 0,67 (17)
Deng 0,71 (32)
Gibson 0,71 (32)
Brewer 0,76 (74)
Rose 0,77 (85)
Noah 0,83 (167)
Boozer 0,95 (340)

Deng's result is even more impressive if we realize how tough defensive assignments he usually has.

My argument last season was flawed. I have come to realize that those numbers only imply either good or bad defense and cannot say X is better than Y because his defensive ppp is 5 point better per 100. Offense is a lot easier to do that with, but you can't with defense. Last season's argument should have been that Rose is a good defensive point guard because he is good everywhere, not that he is better than Rondo, Paul etc. You can't say who is better among them based on that number other than saying they are in the same class of good defensive pgs.

[alucryts]

Raw +/- is a fact, those are real results completely unadjusted. Do you honestly want to imply that those are not facts?
And overall that all has not much to do with "correlation does not imply causation". We are not comparing two unrelated statistical variables here. We have a hypothesis, which can be refuted. We have context and we can test the ability to predict. And so far the predictions seem to confirm the results of that "correlation" while they are not in agreement with your argumentation.

Again, how do you explain that the value of Noah is not seen in the results? If he is so much more important than Deng, the effect should be seen with a different SF too. But it is not. Noah with someone different than Deng at SF creates a worse defense than Deng with someone else than Noah. Why is that, if Noah is the key?

Your right, the plus minus is a fact of the plus minus of a player and nothing else. Any interpretation beyond saying that this is his plus minus ceases to be a fact and begins to be an opinion.

This has everything to do with correlation does not imply causation. Deng's presence correlated strongly with good defense. You are now implying that his presence (the correlation) is the cause of the defense. This is not a leap that is possible without context beyond plus minus. Give me context outside of plus minus that supports your claim.

my argumentation includes the results and the adjusted values.

What are the adjusted values based off of? The results. What are the results based off of? Plus minus. What is plus minus filled with? Insane error/variation. No matter how much correction you apply to plus minus, it will carry with it a large error. You can reduce it to a point where the results can have meaning, but making wild claims along the lines of Deng secretely being the best and most impactful defensive player in the NBA falls into the category of abuse of the stat while completely ignoring uncertainty. You need a much better handle on uncertainty and error before you can start using these stats like you do.

[alucryts]

The biggest problem I have with plus minus arguments is that they are all versions of Russell's teapot.

Russell's teapot, sometimes called the celestial teapot or cosmic teapot, is an analogy first coined by the philosopher Bertrand Russell (1872–1970) to illustrate the idea that the philosophic burden of proof lies upon a person making scientifically unfalsifiable claims rather than shifting the burden of proof to others, specifically in the case of religion. Russell wrote that if he claimed that a teapot were orbiting the Sun somewhere in space between the Earth and Mars, it would be nonsensical for him to expect others not to doubt him on the grounds that they could not prove him wrong. Russell's teapot is still referred to in discussions concerning the existence of God and has drawn some criticism for comparing the unfalsifiability of a teapot to God.

You say that Deng has, for a random made up example, a +4.0 impact on defense ranking him like a DPOY. My argument is that this +4.0 is not only riddled with variance, but it is also in error with correlation and causation. Let's for a second assume 100% accuracy of +4.0 to completely remove variance from the equation. My argument is that +4.0 is not caused by Deng himself entirely. My argument is that a big enough portion of that +4.0 is caused by his team mates and his system to remove him not from being a good defender, but from being the singular MVP of the Bulls (or DPOY). There is no way to prove you wrong under your logic with my opinion because of the plus minus bunker you have put up. Since you are the one here making an extreme claim (Deng is the MVP of the Bulls and DPOY being akin to the teapot in space), the burden of proof is on you. You are not right because no one can provide proof greater than plus minus in your eyes. It is up to you to provide proof beyond plus minus to satisfy your claim. This is the same as saying the person claiming there is a teapot in space is right because no one can prove him wrong. You have insufficient proof.

I also have the burden of proof of showing that Noah is the Bulls best defender. I will attempt to do this with video over spring break for me.

[mysticbb]

Your right, the plus minus is a fact of the plus minus of a player and nothing else.

Do you agree that we can make a prediction on the outcome of certain lineups and compare that prediction with the result in order to determine the quality of the prediction?

Give me context outside of plus minus that supports your claim.

Are you kidding me? I gave the context already, and not only in this thread.

What are the adjusted values based off of? The results.

Yes, lineup results. First step is adjusting for pace and use it as dependent variable, mark the players as independent variables, run the regression.

No matter how much correction you apply to plus minus, it will carry with it a large error.

How "big" do you think the error is? Did you ever done a ridge regression in order to make such bold statement?

You can reduce it to a point where the results can have meaning, but making wild claims along the lines of Deng secretely being the best and most impactful defensive player in the NBA falls into the category of abuse of the stat while completely ignoring uncertainty. You need a much better handle on uncertainty and error before you can start using these stats like you do.

How do you come up with that conclusion? Do you have any evidence besides your personal belief?

The biggest problem I have with plus minus arguments is that they are all versions of Russell's teapot.

I don't think that you have understood the issue brought up by Russell at all. How can you say that the analysis based on regression is unfalsifiable? Because that is the key to Russell's teapot. ;)

There is no way to prove you wrong under your logic with my opinion because of the plus minus bunker you have put up.

That is wrong. You can easily disprove that by using a good enough sample in which your analysis gives a better prediction than the prediction based on RAPM for example.

[alucryts]

Your right, the plus minus is a fact of the plus minus of a player and nothing else.

Do you agree that we can make a prediction on the outcome of certain lineups and compare that prediction with the result in order to determine the quality of the prediction?

Give me context outside of plus minus that supports your claim.

Are you kidding me? I gave the context already, and not only in this thread.

What are the adjusted values based off of? The results.

Yes, lineup results. First step is adjusting for pace and use it as dependent variable, mark the players as independent variables, run the regression.

No matter how much correction you apply to plus minus, it will carry with it a large error.

How "big" do you think the error is? Did you ever done a ridge regression in order to make such bold statement?

You can reduce it to a point where the results can have meaning, but making wild claims along the lines of Deng secretely being the best and most impactful defensive player in the NBA falls into the category of abuse of the stat while completely ignoring uncertainty. You need a much better handle on uncertainty and error before you can start using these stats like you do.

How do you come up with that conclusion? Do you have any evidence besides your personal belief?

Ideally, you can use past to predict future lineups. Unfortunately, in the papers you have provided me in the past, the predictive ability of your stats is terrible. It is the best out there, but it does a terrible job. You provided me this. The paper said that the best predictive version of plus minus has a slight increase in accuracy over simply stating that the home team will win by this many points. If this has changed since then, please provide me the update. If there is none, using these stats as predictors of individual success is ignoring the fact that the error is larger than the result itself.

What is the context outside of plus minus? I have seen none; maybe I have missed it. All of the stats supporting your claim that I have seen are derived from plus minus. Please list the advanced stats that aren’t plus minus, and what they are derived from.

Why is plus minus the only stat or context you count as valid? In your eyes, plus minus is infallible leading to Russell’s Teapot.

[mysticbb]

What is the context outside of plus minus? I have seen none; maybe I have missed it. All of the stats supporting your claim that I have seen are derived from plus minus. Please list the advanced stats that aren’t plus minus, and what they are derived from.

Why is plus minus the only stat or context you count as valid? In your eyes, plus minus is infallible leading to Russell’s Teapot.

I stop here, because the discussion becomes really useless. I provided explantation outside of +/- based analysis regarding Deng's impact MULTIPLE times, and no, I don't need to provide any other advanced stats in order to do so, because the only thing you are bringing up is your personal belief so far with NO data backing it up. So, please, you can't ask me for providing something, when you are the one providing nothing besides a video analysis in which you give out credit based on your personal belief while coming up with the idea that Noah is the real defensive MVP of the Bulls. You can't explain a tiny little bit, why that is not seen in the results. The pair Noah/Korver for example is worse defensively than the pair Korver/Asik, Korver/Thomas or Korver/Dalembert for the last 4 years (2008 to 2011). How would you explain that?

You don't understand ridge regression at all, otherwise you wouldn't ask for standard errors of that analysis. You don't trust that, because the RMSE (root mean square, that is NOT the standard error) is too big for you (talking about the paper I provided), without understanding that this error nees to be applied to the result (ORTG and DRTG, which is about 105 for that test), and not to the difference (do you even understand that in average the percantage of the standard error wouldn't even be defined, if it would have been used in the way you are using it?). HCA is a WORSE predictor, so, don't act like it would be the same.

And to claim that the hypothesis would be unfalsifiable, because you are unable to prove me wrong, is hilarious. It is the other way around here, you are making claims without any way to prove or disprove that. You just expect that we believe in your method and ability to analysis that correctly. You are giving here the Russell's teapot. ;)

[alucryts]

Like I said, I also have the burden of proof of proving my theory. I have already admitted this.

The data that you believe proves your point in fact does not prove what you think it proves at all. Your entire argument here is that because I cannot provide data to contradict yours, you must be correct. You are also incorrect in assuming that individual defensive impact is capture-able and separable from the rest of the team. There is no way to separate player from system, team, and lineups to come up with a truly individual stat that can stand on its own. Even the best of these stats is still laced and overwhelmed with synergistic relationships on the defensive end. If one player continually benefits more from the synergistic relationships more than the other, does that make him a better defender, or does it make him a good defender in a good situation? That is really the question here.

The reality of the situation is that Rose depends on Noah who depends on Deng who depends on Boozer who depends on Rip. Now take that web and interconnect it for every possible connection between five players on the floor at any one time. There is no way to pull that web apart to a high enough degree to really come out with an outlandish claim that Deng is the DPOY, MVP of the Bulls, and the reason for the Bulls success more than any other player or coach. Deng would likely have to be one the best defensive players of all time in order to pull off all of that.

You have a very extreme claim backed by a flawed stat and are unable to provide enough proof to back it up. On top of this, you are unable to recognize the flaw in the stat and compensate for it with more solid evidence. Plus minus and its variants and cross correlations are not sufficient proof. There is an inherent degree of inaccuracy in any advanced number, and all evidence points to Deng’s being inflated to a level that is not realistically achievable by a player in his position.

[mysticbb]

The data that you believe proves your point in fact does not prove what you think it proves at all.

I use the REAL results as proof for that. So, what should be used instead?

Your entire argument here is that because I cannot provide data to contradict yours, you must be correct.

Correct, I'm correct as long as you can't provide data which disproves that. That is the way science works. ;)

You are also incorrect in assuming that individual defensive impact is capture-able and separable from the rest of the team.

Why? Because you are saying it?

Every game snippet can be used to make that seperation. If player x with the same teammates has a better result than player y while adjusting for the strength of the opponents, it might be the case that player x is doing something on the court which helps more. For Deng it is his defensive abilities, for others it might their offense. We can make that seperation to a certain degree for all players, it becomes better with increased sample. For RAPM that "better" is seen in an increased standard deviation of all values, because it just pulls everything closer to the mean otherwise. That's why uninformed RAPM has lower values than prior informed, that's why a full season dataset will give bigger numbers for the top players than a dataset based on half season. That is how that works. The more data, the better the seperation.

If one player continually benefits more from the synergistic relationships more than the other, does that make him a better defender, or does it make him a good defender in a good situation? That is really the question here.

For the context of a 5on5 game, the player who makes his team playing better defensively, might be as well be the more valuable defensive player. If the player is not that good defensively, it should create contradictions within the dataset or should show up in the results.
No player can play minutes like Deng and end up being wrongly assigned to positive value while it should have been on another player. Seriously, we DON'T have a case in which a small minute roleplayer has a big value and the result is basically just caused by overfitting (well, given the fact that I use ridge regression, overfitting can't be an issue at all). If we have huge multicollinearities, both related players would get pretty much the same value assigned too. That is the result of ridge regression here instead of normal regression.

The reality of the situation is that Rose depends on Noah who depends on Deng who depends on Boozer who depends on Rip. Now take that web and interconnect it for every possible connection between five players on the floor at any one time.

Nobody is disputing that. But in that sense, you can't say anything about any individual player at all anymore while using results as justification. With your argumentation you just rendereed game results useless.

There is no way to pull that web apart to a high enough degree

What makes you say that? Do you have any further argument to support your claim here?

to really come out with an outlandish claim that Deng is the DPOY, MVP of the Bulls, and the reason for the Bulls success more than any other player or coach. Deng would likely have to be one the best defensive players of all time in order to pull off all of that.

Sorry, but that is complete nonsense. The Bulls without Deng are playing at a pretty good level. He does not need to be the best defensive player ever in order to be MVP of the Bulls or DPOY. Was Howard last season the best defensive player ever?

You have a very extreme claim backed by a flawed stat and are unable to provide enough proof to back it up.

I used the FREAKING RESULTS! What else should I use? Seriously, give me a thing which i should use in order to provide MORE hard evidence.

And believe me, I know that the stats has flaws. I would in no way claim that this is completely accurate. But so far I don't see anything provided by you which make it seem that you are more accurate. In fact, so far we have seen NO evidence at all from you.

[drza]

What is the context outside of plus minus? I have seen none; maybe I have missed it. All of the stats supporting your claim that I have seen are derived from plus minus. Please list the advanced stats that aren’t plus minus, and what they are derived from.

Why is plus minus the only stat or context you count as valid? In your eyes, plus minus is infallible leading to Russell’s Teapot.

As someone who has been following this exchange with interest, I would say that both sides have provided context and analysis outside of quantitative analysis. Actually, in several posts. Alucryts had the excellent post where he broke down the three main epochs of the Bulls defense, and others where he's given his opinion on how big men play a larger role in team defense because they are usually the ones that have help responsibilities when other players funnel their men into the bigs. This is the traditional view of team defense, and has been supported well through the years. In the decade since we have access to +/- stats, though, there are several small forwards that have shown up with defensive impacts on the order of the best big men...there are a lot MORE big men on that level, but some small forwards do show up. But I'm going to stop with the +/- talk for now, because I'm making a feeble attempt at mediating and as yet +/- results have not been fully accepted by both sides. Fair enough.

Mysticbb's contextual/analytic contributions outside of +/- have had to do with a combination of Deng's ability to play good iso D, his ability to funnel/steer his opponent towards the help better than usual, and his ability to make good defensive rotations with very few mistakes. Alucryts, you may not agree with mysticbb's points, but they HAVE been stated several times in this thread. It hasn't all been +/-.

So then, to this point, we are at a bit of a stalemate. Alucryts would call Noah the defensive MVP because, in Thibs' defense, Noah is playing the mobile/help big man role which has a large responsibility. This player is a central part of the defense by definition, because he has to be able to provide help from the rim on out to wherever a screen is set and he always has a host of responsibilities. Alucryts has also said in this thread that Asik and Gibson working together roughly approximate what Noah is able to do defensively by himself. My interpretation is that Mystic, on the other hand, would argue that though Noah's ROLE is very important to the defensive set, that how well Noah FULFILLS that role is in question. Noah's role likely is more important than Deng's role, but if Deng does his job on defense enough BETTER than Noah does his job, Deng could still be the more impactful defender.

At least, that's my interpretation of what I've read from the two of you, completely outside of what +/- might say. And at this point, the debate reaches good barber-shop level but can't really go much beyond because there's no way to prove either side correct based purely on context. And THAT is where the numbers come in, because we can make some logical conclusions based on the results. Alucryts, you reasonably bring up the burden of proof and the difference between correlation and causation. But on the flip side, the +/- numbers aren't the case themselves, they are the SUPPORT of mystic's case as laid out above.

In other words, both of you have a case for what you think is happening on the Bulls' defense. Mystic has been using +/- and the team results with different line-ups to support his case. At this point, Alucryts, I'd say the burden now shifts to you to provide some type of quantification to support your own interpretation. You can argue that how mystic has interpreted the numbers is incorrect (which you have), but what you haven't done is offered an alternative analysis of the numbers that fits your own case.

For instance, alucryts says that Boozer is the weak link that hurts Noah's individual +/- ranking. Mystic responds by pointing out that even adjusting Boozer to be an overly awful defender doesn't improve Noah's individual rating to be higher than Deng's. In the absence of further discussion, to me this point would seem to go to mystic.

Another example is that alucryts says that Korver is such a terrible defender that he is the one biasing the results. Mystic responds by pointing out that the Korver/Noah duo has worse results than Noah with 4 other centers over the past several years, which would at least suggest that the entirety of the results can't be placed on Korver.

But in the face of this type of quantitative data point, alucryts' response thus far has been essentially that you can't make any kind of conclusion based on +/- data. That defense is too interrelated to separate out the contributions in a meaningful way. As an experimental scientist by profession I would strenuously disagree with that assertion. Making conclusions based upon experimental observation is at the heart of the entire scientific method, and I'd dare say that arguing against that method of analysis being useful would by far be the more "Russell's teacup" stance to take in this debate. No, the question isn't whether it's POSSIBLE to make a meaningful conclusion based on observation...the question is HOW GOOD OF A JOB the current +/- tools we have available do it, and how many observations under how many conditions are required before those results become meaningful. And THAT is a very different statement, on the burden of proof scale.

Because there is definitely some degree of support that +/- stats are in some way providing useful data. There's the more peer-reviewed-scientific support of the order of what will be displayed at the MIT Sloan conference, but short of getting that deep there is also the anecdotal support that so many NBA teams are beefing up their stats department and making more decisions based on these +/- results. My point in this paragraph isn't to rigorously prove that +/- data is useful, as much as to say that there IS very clear, very definite support that +/- data is useful. As such, the "Russell's teacup" defense doesn't work here, allucryts, because the initial premise that +/- data can be useful in player evaluation is NOT a nonsensical claim on the order of a teacup orbiting the sun.

Said another way, it's like this is a legal case and one side is making the argument that the other side's case is without merit and therefore doesn't have to be argued. But with the body of +/- work out there and it's growing adoption into the NBA decision-making lexicon, you can't just stand on a "without merit" claim alone. You actually have to try your case, which means that simply ignoring the +/- evidence that mystic provides or claiming it's not relevant doesn't make your case any stronger. Like DocMJ, I am very interested in the type of video-based analysis you were providing. But in order for it to be believable, considering all of the quantitative data we have available, I would need to see some type of numerical/results-based evidence to support your visual analysis. Otherwise, it just boils down to your opinion vs. someone else's, and while an educated opinion is cool in the end it's ultimately not very convincing. At least to me.

[mysticbb]

drza, thanks for that well written post. You covered my part competely! And I understand alucryts in the same way as you are doing it.

[Dr Positivity]

I think you could make the analogy that Noah is like the hockey goalie and Deng is the defensemen. The goalie is always the most important player on the team systematically (it's not even close). But defensemen can have more value - and do more often than not - because what matters is vs replacement

(In this case though I *do* think that getting a C that's good defensively like Noah is really really hard. It just so happens that the Bulls have a 2nd one in Asik)