What's the best defensive metric

[EvanZ]

On what basis do you make that assertion? It seems you've come up with an ad hoc assignment of blame based on gross generalization ("defense at the point guard position is a luxury").

It's based on looking at the minimum and maximum defensive RAPM at each position last season:

PG -2.9...2.2
SG -1.9...3.3
SF -1.9...3.1
PF -2.2...3.7
C -2.8...3.1

Note that the low end of the PG (-2.9) is almost a full point lower than every other position, except center (we'll get to that in a minute). And the maximum DRAPM for PG is almost a full point lower than every other position. That's what I'm referring to.

By the way, for reference, Calderon's DRAPM last season was -0.2.

And which center was -2.8? Ok, that's easy. It was Bargnani. Take him out, and the lowest DRAPM for centers last season was -2.2.

I understand you think Calderon is horrible. Try watching Stephen Curry, or maybe Luke Ridnour for a full season. Like I said, I don't think Calderon's defense is your biggest problem right now. Because you ain't gonna be able to replace him with somebody much better in all likelihood. But Bargs? Pretty much any stiff would be better. Hell, even Andris Biedrins would be an improvement on defense.

[floppymoose]

Just look at raw +-. Bargs had the worst defensive raw +- on the team. It's difficult for a player to do that if he isn't the worst defender on the team, at least of the big minute players. Calderon's raw defensive split was not nearly as bad.

It's not like RAPM is coming up with some mysterious answer here that isn't obvious just from looking at the raw numbers. My eyes tell me Bargs is the biggest problem. The score tells me Bargs is the biggest problem. And PDSS is telling me Bargs is like his teammates. One of these things is not like the other... one of these things does not belong...

[Wannabe MEP]

It's based on looking at the minimum and maximum defensive RAPM at each position last season:

PG -2.9...2.2
SG -1.9...3.3
SF -1.9...3.1
PF -2.2...3.7
C -2.8...3.1

This is much more pronounced if you look at multi-year RAPM, because RAPM improves with more data.
The top defenders according to multi-year RAPM:
6.2 KG
4.6 Bogut
4.1 Duncan
4.0 Collison
3.9 Luol Deng
3.9 Tony Allen
3.8 Howard
3.6 Chuck Hayes
3.6 Josh Smith
3.6 Przybilla
3.6 Jason Collins
3.5 Jermaine O'neal
3.4 Jarron Collins
3.3 Amir Johnson
3.2 Camby

Almost entirely bigs--zero pgs: I'd have to keep the list going for quite a while before I found any pgs. The difference between a great defense big and a mediocre defensive big is enormous...the difference between a great defensive big and a sh*tty defensive big like Bargs is superlatively enormous.

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.
2) Bigs live in the paint, where all the action is. A mobile, long, tenacious big patrolling the paint makes easy buckets almost impossible, which makes all the difference in the world (See: Russell, Bill).
3) The NBA is pick-and-roll basketball. The big is the most important defender against the pick-and-roll: an alert, quick big can hedge to stop penetration and still recover to prevent an easy bucket for the screener. Prime Tim Duncan is the #1 reason why the Nash-Stat pick-and-roll never won an NBA Championship, or even made an NBA finals.

[EvanZ]

Good stuff. I especially like yoir first point. It's so obvious in retrospect, but I've never really thought about it that way.

[Paydro70]

It's also patently false. If it were true, teams would be composed of entirely centers.

[EvanZ]

It's also patently false. If it were true, teams would be composed of entirely centers.

If centers could take the ball up the floor, maybe they would be.

[floppymoose]

Absolutely. When you have height that can play the guard roles well, you use it. None of the regular starters for the 96-97 Bulls were under 6'6", and they didn't do badly as a team.

[Wannabe MEP]

It's also patently false. If it were true, teams would be composed of entirely centers.

Apparently it's news to you that there are two parts to basketball?!?

We're specifically talking about defense. All we've said is that a team's most important defenders are it's bigs. On offense, the most valuable players can easily be the shortest guy on the court, whose height is not really a problem on the defensive end if there's good defensive length in the paint. The best team possible is going to be composed of a quick, crafty, unselfish facilitator combined with good complementary offensive players (shooters/slashers) and a great defensive frontcourt.

The facilitator doesn't have to be the PG, although that makes the most sense:

1) It's logical that your best dribbler is your facilitator and the dude bringing the ball up the court
2) Lowest to the ground = lower dribble = less opportunity for steals
3) Lower center of gravity + less mass = increased agility (that's pure physics, not to mention physiology)
4) ??? There may be something to the idea that a quick, crafty short guy disappears in traffic like Messi in futbol ???

Offense is much more complicated than defense, but multi-year RAPM still has some pretty clear trends. The best offensive RAPM scores shade toward players who are:

1) Elite facilitators
2) Excellent shooters
3) Quick for their position

8.2 Steve Nash
6.4 Dwyane Wade
5.9 Dirk Nowitzki
5.3 Manu Ginobili
5.3 LeBron James
5.1 Chris Paul
4.9 Kobe Bryant
4.6 Baron Davis
4.1 Chauncey Billups
3.5 Kevin Durant
3.5 Deron Williams
3.3 Jamal Crawford
3.2 Ray Allen
3.2 Carlos Boozer
3.0 Jason Terry
3.0 Carmelo Anthony
3.0 Kevin Martin
2.9 J.R. Smith
2.8 Jason Richardson
2.8 Antawn Jamison
2.8 David Lee
2.7 Russell Westbrook
2.7 Devin Harris
2.7 Joe Johnson
2.6 Paul Pierce
2.6 Danny Granger
2.5 Ty Lawson
2.5 Andre Miller
2.5 Mike Conley
2.4 Stephen Curry

PGs: 11
8.2 Steve Nash
5.1 Chris Paul
4.6 Baron Davis
4.1 Chauncey Billups
3.5 Deron Williams
2.7 Russell Westbrook
2.7 Devin Harris
2.5 Ty Lawson
2.5 Andre Miller
2.5 Mike Conley
2.4 Stephen Curry

SGs: 10 (many who function as facilitator)
6.4 Dwyane Wade
5.3 Manu Ginobili
4.9 Kobe Bryant
3.3 Jamal Crawford
3.2 Ray Allen
3.0 Jason Terry
3.0 Kevin Martin
2.9 J.R. Smith
2.8 Jason Richardson
2.7 Joe Johnson

Hybrid forwards: 9
5.9 Dirk Nowitzki
5.3 LeBron James
3.5 Kevin Durant
3.2 Carlos Boozer
3.0 Carmelo Anthony
2.8 Antawn Jamison
2.8 David Lee
2.6 Paul Pierce
2.6 Danny Granger

Centers: 0 (unless we want to count David Lee here?? I mean, he's undersized for a PF...)


Gee, so now teams should be composed entirely of guards, right?

[giberish]

Gee, so now teams should be composed entirely of guards, right?

Don Nelson says Hi.

[turk3d]

Great thread. Excellent points all around and puts a lot of these metrics in perspective which is what I think is needed. And special props to BorisDK1 for all the work and your thorough analysis and to Los Soles for your clear unbiased approach. It's all about coming up with the best you can right, guys. Keep the discussion ongoing, you are all making good points, I think it's great.

I really like Boris approach, pure stats which are based on raw data are ok (especially when that's all you have, they're better than nothing), but they can't fully take the place of what actually happens on the court, which is what I think Boris is actually charting. A lot of work, that's for sure. I do think that this is the reason we have to always look at the games in conjunction with stats before coming to any final conclusions about a player, his worth and how he impacts the team.

Once again, great thread, I appreciate all your comments. I have a tendency to not really delve all that much into these advanced stats, in part because they're limited in terms of their availability and also because of the conclusions that some who promote them come up with which just don't always jibe with what I obesere on the court.

There are just too many other factors (due to the games complexity which you realize if you've been around the game for a long time, it's no longer just putting a basket in a hoop any more) which don't get taken into account in the computations being made. As we know, there are always more complexity that can be added to the formulas which are used.

This is how engineering typically works and we know that Engineering in general is usually a series of approximations. It's just a matter of how many iterations you feel you need in order to get it to where by adding another degree of granularity gets you diminishing returns. Most of these formulas being used are approximation and so the question is do they go out far enough? I think the answer may be no. And it sounds to me like Boris has taken it to the next level, but only for his team.

The question then I ask to any of you who promote using advanced stats or even working on them, are you doing this to promote your agenda (if you have one) or are you doing this to come up with the best, most unbiased amd accurate one you can possibly come up with, which is what I think should be the overall goal.

[Amen316]

I used DeanO's Project Defensive Score Sheet as the basis for tracking the Wizards defense over the course of about 3 seasons. I collected additional information as well, then later worked with Dean on something that basically put PDSS on steroids. I understand floppymoose's concern, but the information was robust. I actually worked with a few teams providing them info from this kind of tracking. Front offices usually liked the data, and some coaches REALLY liked it. A few coaches didn't like it and said they wouldn't use it. Worth mentioning that comments like that invariably came from coaches who disagreed with the numbers about a particular player or two.

Anyway, I never got traction beyond 5-10 game stints because teams traditionally don't pay very much for statistical analysis. So many people want to "in" that they're willing to virtually give away their stuff. Teams weren't willing to pay me what I would need to be willing to do the tracking over a long period of time. (And it wasn't like I was asking for a max contract either. )

The real value of PDSS is NOT being able to generate a defensive rating. The real value came when we put thing on 'roids and were tracking by offensive play types, help defense, ball movement, and an array of other stuff. Being able to quantify who was good at what was great information to provide for coaches/front office types. Another value from systematic tracking: it measures effectiveness. Sometimes a player may look like he's doing a good job on film, when he isn't. That showed up in PDSS. Other times, there are guys who don't "look good," but are actually quite effective when it comes to the goal of defense: preventing the other team from scoring. That also showed up.

In many ways, I understand the coaches who said they wouldn't use it. They're typically inundated with advice. Everyone KNOWS what the coach should be doing. But a lot of the advice they get is bad. So ultimately, the coach has to trust himself and his staff.

I like Synergy and find their info useful. It's good enough (and cheap enough) that it pushes aside PDSS -- at least at the team level. Front office types think they have basically the same info they'd get from PDSS from Synergy. I don't think Synergy is as good, but I could be biased. I wish Synergy would adjust use terminology consistent with APBRmetric lingo. What they call a "possession" is what APBRmetrics would call a "play." As possession in APBRmetric parlance lasts until the other team gets the ball back. In Synergy, an offensive rebound creates a new possession. In APBRmetrics, an offensive rebound is a play within a possession -- it extends the possession, it doesn't start a new one.

Great Post
( I too have participated in stat info and well over 50 transactions with NBA teams and your correct 90pct of the time there is no money involved even when you move up from statistical info to actual consulting. Trades - Contracts - Stats - Draft - Free Agent all part of their daily work behind the scene.) Synergy is helpful but falls way way too short of its ability simply because the people inputting the info have turned into basic robots and are missing so so many moments of info they could be inserting into equations. In the end you find yourself going back through the robotic work that Synergy is suppose to be supplying. I spoke Synergy about this but in the end our conversations ended in advise from their lawyer that I could be talking about things they want to do in the future and communication stopped.
As far as APBR and possession if you ask 90 pct of the coaches out there they will agree that every rebound is a new possession. APBR metric is totally off based here because the moment a shot leaves a shooters hand it is no longer in a teams possession. That ball becomes neutralized (possession is over) in flight and either it goes in or it is rebounded by one of the teams to create a new opportunity.
That is fact this isnt like a pass from one player to the next control is lost and outcome becomes pct based.)

[bijoy12]

Of the defensive metrics out there, the Fielding Bible's +/- is considered the most rigorous (but it doesn't rate catchers and pitchers and it's got some major problems with its' fundamental assumptions). After than you've the zone based UZR/RZR/PMR group which have the best info incorporated into their algorithms but are still based on a bad initial assumption that defense is an individual effort and that fielders stay within set zones at all times. After that you have team based defensive metrics like DWS and PCA fielding which are based on better assumptions but thus far do not incorporate the best data and below all of those are the now completely outmoded Davenport Fielding Translations and associated FRAA, FRAR and RATE family from BP and then the basic fielding metrics (ZR, RF and F%).
san antonio dentists

[floppymoose]

^^^ lol, nice spambot.

[Paydro70]

It's also patently false. If it were true, teams would be composed of entirely centers.

Apparently it's news to you that there are two parts to basketball?!?

We're specifically talking about defense. All we've said is that a team's most important defenders are it's bigs.

No, the original quote, the one I responded to, was "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."

This is obviously false. A center on a point guard is a mismatch, the reality of the human body makes it impossible for a much taller player to stay in front of a much shorter one (who is also likely to be a better athlete). We can quibble about what the word "roughly equivalent to" means, but we've seen countless examples in the past of players who are successful defenders despite giving up 4" to their matchup. Just as an anecdote, Bird always said Michael Cooper was the toughest defender he faced.

That's also only talking about the half-court... obviously open-court play favors shorter players.

[EvanZ]

Centers regularly defend against PG's by blocking their shots. (Or is that not considered defending?)

[turk3d]

Centers regularly defend against PG's by blocking their shots. (Or is that not considered defending?)

Although true, what you're really referring to is "help defenders" and the truth is anyone can do that if they want to and typically when a Center winds up on a guard (or any defender winds up defending someone other than his man) he's left the man who he's basically responsible for and is instead helping one of his teammates (who's been beaten by his man or who's in a decided mismatch). However, it's just not his main responsibility (as if in the case of a Center) he were defending another Center, or PG defending PG, etc., etc.

[EvanZ]

The bottom line is that centers regularly block shots. Point guards do not, and most cannot.

And I'm not sure if you know this, being a Warriors fan and all, but "help defense" is kind of a big deal.

[Paydro70]

I have no idea what that's the bottom line to. The original claim was that centers can defend every position... they can't, even in the half court, and basketball is not exclusively played in the half court.

That they are the most important defensive position does not mean that their defensive skillsets are universal.

[turk3d]

The bottom line is that centers regularly block shots. Point guards do not, and most cannot.

And I'm not sure if you know this, being a Warriors fan and all, but "help defense" is kind of a big deal.

It's a very big deal, not denying that. Just pointing out that's it's a different facet (in comparison to playing against your assigned man) which is what we think of when we look at a man's defensive prowess. It's essentially a double-team versus an individual matchup but yes, perhaps a good big man's biggest impact will be on the defensive end where he helps his teammates (same can be said for others, ie small guys who can get steals and block the passing lanes).

That's why I don't (like many others) deny the value of players who get a lot of steals as being part of defense and minimize their value. Much of the time that player (who's getting steals or deflections) is providing help defense to one or more of his teammates. Even when he doesn't get credit the steal (which is what gets shown on the stat sheet) he may be saving another player from getting burned or stopping a fast break and easy bucket from being scored by the other team.

[EvanZ]

I have no idea what that's the bottom line to. The original claim was that centers can defend every position... they can't, even in the half court, and basketball is not exclusively played in the half court.

Centers are the last line of defense against *everyone* driving to the basket. Point guards are the first line of defense against other point guards and sometimes shooting guards.