What's the best defensive metric

[floppymoose]

Centers end up augmenting the defense, against everyone, much more than point guards do. I think that is what is being talked about here. I'll link this article again:

http://www.backpicks.com/2011/01/22/defensive-usage/

[Wannabe MEP]

Paydro, really?!? Belligerently dense is common on these boards, but not from the mods.

you reply to a single quote in a long and involved post in a long and involved thread giving no indication that you were referring to that single quote until much later. you rip the quote from context, misrepresent it, and completely ignore the general purpose of the original post. you spew trash about it with no argument, no defense: just your arrogantly stated opinion. You make an obviously false assessment that ignores every other aspect of basketball, instead of directly challenging the point in question. and you do zero analysis on any of the statistics in a statistical discussion in the statistics forum.

the whole point is:

That they are the most important defensive position

great. we agree. we also agree that we're not going assign our center to defend the other team's point guard.

[Paydro70]

Paydro, really?!? Belligerently dense is common on these boards, but not from the mods.

I have no idea which amongst my comments could be called belligerent. Someone said something I didn't think was true, which another poster then agreed with. I thought the statement was so self-evidently false that I first responded with a flippant hyperbole. I'm sorry if I offended you.

you reply to a single quote in a long and involved post in a long and involved thread giving no indication that you were referring to that single quote until much later.

My first post in this thread came directly after EvanZ's post: "I especially like yoir first point. It's so obvious in retrospect, but I've never really thought about it that way."

The "first point" was: "1) You can only truly defend people that you're taller than or roughly equivalent to. Centers can defend everyone on the court, while pgs defend pgs."

Not only do I not think the point is "obvious," but I don't think it's true, and in my second post I explained why.

you rip the quote from context, misrepresent it, and completely ignore the general purpose of the original post.

...what is the "context" missing here? If I might paraphrase the post, it said "bigs are by far the most important players defensively. Here are some reasons why." What about that context changes the fact that I think one of those reasons is false? What does the general purpose of the original post matter in this case?

If the original post were a diatribe about how Shaq was the greatest player in NBA history, and one of the reasons given was that he could defend point guards, I think it would be reasonable to "rip from context" that particular reason and disagree with it, no?

you spew trash about it with no argument, no defense: just your arrogantly stated opinion.

True, in my first post I gave only that flippant comment about teams being entirely centers. But in my second post, I think I made very clear what I was talking about and the basis for my thinking. Maybe you were not convinced, but I think the logic is fairly sound, since you make much the same point that I did in one of your own posts: "3) Lower center of gravity + less mass = increased agility (that's pure physics, not to mention physiology)"

In other words, taller players have certain disadvantages vis a vis shorter players, even on defense, and therefore centers cannot "defend everyone on the court," nor is it impossible for a shorter player to defend a taller one.

You make an obviously false assessment that ignores every other aspect of basketball, instead of directly challenging the point in question.

I assume you mean my "only centers" comment? I'll apologize again, then.

and you do zero analysis on any of the statistics in a statistical discussion in the statistics forum.

I have tried to make a point about one of your explanations for a particular statistical trend. I think that qualifies as a legitimate post for the forum.

[Nivek]

I think you guys are talking past each other. Los Soles and EvanZ are making the obviously accurate point that centers participate in defending every position on the floor. They do this through help defense and by typically being the last line of defense at the rim.

Paydro is making an obviously accurate point that centers are not and should not be matched up (assigned one-on-one defensive responsibilities) against an opposing PG.

Of these, I think the first point is more relevant to the discussion we're having. We typically make a big deal about individual matchups in analysis, but close tracking and study of how games are actually played reveals that one-on-one matchups aren't what matter. While individual talents do matter a great deal, basketball is still a team game. One team works together to score; the other team works together to prevent a score.

Within the context of that defensive teamwork, a center plays the most important role. First, as the defender closest to the basket, which is where shooting percentages are highest. Second, as a help defender. Third, as the tallest defender, which means he's likely to lower opposing shooting percentages more effectively than his smaller defensive teammates.

[turk3d]

I think you guys are talking past each other. Los Soles and EvanZ are making the obviously accurate point that centers participate in defending every position on the floor. They do this through help defense and by typically being the last line of defense at the rim.

Paydro is making an obviously accurate point that centers are not and should not be matched up (assigned one-on-one defensive responsibilities) against an opposing PG.

Of these, I think the first point is more relevant to the discussion we're having. We typically make a big deal about individual matchups in analysis, but close tracking and study of how games are actually played reveals that one-on-one matchups aren't what matter. While individual talents do matter a great deal, basketball is still a team game. One team works together to score; the other team works together to prevent a score.

Within the context of that defensive teamwork, a center plays the most important role. First, as the defender closest to the basket, which is where shooting percentages are highest. Second, as a help defender. Third, as the tallest defender, which means he's likely to lower opposing shooting percentages more effectively than his smaller defensive teammates.

Yes, you're right in all your points however, is that the question that the OP wanted answered? I think not. If the question were: What's the most important position defensively on the court, I think that hands down, you'd get Center for answer but the real question I believe is what's the best defensive metric? And I think that's referring to every position (otherwise it would need to be rephrased to say "What's the best defensive metric for the Center position?).

So either you could give a definition for each position (they could be different if given some thought) or you could try and come up with an overall definition (probably would be more difficult to come up with a fair comparison) because defense is likely position dependent as it is also probably defensive system dependent (which may vary from team to team).

I'm sure that is why the question was asked in the first place and why it's rather difficult to come up with a fair system of measurement for measuring a player's defensive production. Needless to say it's unfair to try and compare the defense of a PG to the defense of a Center. I think that maybe one of the things that Evan has worked on has taken this into account (he may use some sort of a weighted average in his computations)>

[Nivek]

I think you guys are talking past each other. Los Soles and EvanZ are making the obviously accurate point that centers participate in defending every position on the floor. They do this through help defense and by typically being the last line of defense at the rim.

Paydro is making an obviously accurate point that centers are not and should not be matched up (assigned one-on-one defensive responsibilities) against an opposing PG.

Of these, I think the first point is more relevant to the discussion we're having. We typically make a big deal about individual matchups in analysis, but close tracking and study of how games are actually played reveals that one-on-one matchups aren't what matter. While individual talents do matter a great deal, basketball is still a team game. One team works together to score; the other team works together to prevent a score.

Within the context of that defensive teamwork, a center plays the most important role. First, as the defender closest to the basket, which is where shooting percentages are highest. Second, as a help defender. Third, as the tallest defender, which means he's likely to lower opposing shooting percentages more effectively than his smaller defensive teammates.

Yes, you're right in all your points however, is that the question that the OP wanted answered? I think not.

I've already addressed the OP's question -- back several pages in this thread. I was addressing (and attempting to resolve) a conversation that has carried on for the past couple pages, which has not much to do with the original question posed at the beginning of this thread.

If the question were: What's the most important position defensively on the court, I think that hands down, you'd get Center for answer but the real question I believe is what's the best defensive metric? And I think that's referring to every position (otherwise it would need to be rephrased to say "What's the best defensive metric for the Center position?).

As I've written, I think the best metric is Project Defensive Score Sheet on Steroids. It does a good job of capturing the contributions of all defenders.

So either you could give a definition for each position (they could be different if given some thought) or you could try and come up with an overall definition (probably would be more difficult to come up with a fair comparison) because defense is likely position dependent as it is also probably defensive system dependent (which may vary from team to team).

An "overall" definition is basically a waste of time because of differing responsibilities. It's not 5-position independence, however. There's overlap between PG/SG; SG/SF; SF/PF; and PF/C. Responsibilities can probably be broken down into a couple broad categories -- perimeter defenders and bigs. Although, responsibilities and assignments can vary significantly here too. This is one advantage of Project Defensive Score Sheet Steroids version -- it tracks offensive play type, help defenders, and other stuff. I was tracking penetrations allowed, for example. Recoveries to perimeter shooters -- that sort of thing.

I'm sure that is why the question was asked in the first place and why it's rather difficult to come up with a fair system of measurement for measuring a player's defensive production. Needless to say it's unfair to try and compare the defense of a PG to the defense of a Center. I think that maybe one of the things that Evan has worked on has taken this into account (he may use some sort of a weighted average in his computations)>

While responsibilities differ, a PG's defense can be compared to a Center's by looking at how good a job a PG does at a PG's responsibilities vs. how well a center does at performing a center's defensive responsibilities. That said, an average center may still end up with a greater defensive impact/effectiveness than a top PG. That's because of the center's size (height matters when defending opposition shot attempts) and his role.

[Wannabe MEP]

I think you guys are talking past each other. Los Soles and EvanZ are making the obviously accurate point that centers participate in defending every position on the floor. They do this through help defense and by typically being the last line of defense at the rim.

Paydro is making an obviously accurate point that centers are not and should not be matched up (assigned one-on-one defensive responsibilities) against an opposing PG.

Of these, I think the first point is more relevant to the discussion we're having. We typically make a big deal about individual matchups in analysis, but close tracking and study of how games are actually played reveals that one-on-one matchups aren't what matter. While individual talents do matter a great deal, basketball is still a team game. One team works together to score; the other team works together to prevent a score.

Within the context of that defensive teamwork, a center plays the most important role. First, as the defender closest to the basket, which is where shooting percentages are highest. Second, as a help defender. Third, as the tallest defender, which means he's likely to lower opposing shooting percentages more effectively than his smaller defensive teammates.

Yep, I agree with all that.

[Wannabe MEP]

Yes, you're right in all your points however, is that the question that the OP wanted answered? I think not. If the question were: What's the most important position defensively on the court, I think that hands down, you'd get Center for answer but the real question I believe is what's the best defensive metric?

A bit removed from the original discussion perhaps, but it's definitely relevant:

What's the best defensive metric...To help evaluate a player's defensive worth to a team? I understand metrics aren't the silver bullet to establish one player better than the other, but much like PER is to offense, a nice guideline, is there a similar statistic/metric to evaluate an individual's defense? Thanks

In answering the original question, RAPM and PDSS became leading candidates. The discussion about centers was part of a larger debate between PDSS and RAPM. My post was a defense of the RAPM findings and how they are superior to Boris's PDSS findings of the Raptors. RAPM and Boris's PDSS findings came to dramatically different conclusions about the Raptors defensive issues. This is still a massive, unresolved issue in this discussion. I was extrapolating about what RAPM says and explaining why I believe it says that.

The point I'm still trying to make is: The answer to the original question is multi-year RAPM/APM.

The appropriate caveat is that PDSS is incredibly valuable for more specific analysis, BUT IT IS NOWHERE NEAR AS GOOD AS AN OVERALL EVALUATIVE NUMBER (like PER is, which is what the OP was looking for).

[Wannabe MEP]

And Paydro, I think we both took it too far; sorry about my part in that; it seems that we generally agree; moving on.