RealGM

I may live to regret this shocking headline, but with Bryant now tying the record for most All-Defensive 1st team awards, we have a problem.

I'd encourage you to read the whole piece but here's a chart from it, where I've taken the 20 players in recent basketball who come to mind as celebrated and sorted them both by All-Defensive accolades and APM scores.

Bottom line: Kobe routine scores as a defender of average impact by these metrics, despite being a lock to get the All-Defensive accolades every year. Something ain't right.



http://asubstituteforwar.com/2011/05/09 ... maginable/

Well there goes my next blog post.

I mean, maybe Kobe will be 1st-team all-D after he retires. Maybe in 2050 he'll still be on the team. It's bordering on the most bizarre accolade-driven lunacy I've seen in sports in my life. Maybe ever.

ElGee wrote:Well there goes my next blog post.

I mean, maybe Kobe will be 1st-team all-D after he retires. Maybe in 2050 he'll still be on the team. It's bordering on the most bizarre accolade-driven lunacy I've seen in sports in my life. Maybe ever.

lol

Yeah, I've totally cemented my rep as a lunative Kobe hater now, but the scale of the issue here is massive. We're literally talking about an average-on-average defender getting named one of the top 5 defenders in a 400+ player league as many times as any other player in history now. If being the most decorated in history still doesn't give any certainty that you're excelling in the area in question, that accolade is rendered almost completely meaningless.

People say stuff about an award becoming meaningless all the time and they go way too far. As TrueLAfan says about the MVP: You may not agree with who one, but the shortlist is almost always correct. Controversies involve a bunch of guys above the 99th percentile in the NBA vying for an accolade that anoints them to the 99.5th percentile.

What we're talking about here is a guy at the 50th percentile being anointed to the 99th percentile as a given each year. Utterly mind blowing.

In the past there was at least the argument Bryant could turn it on, but the series against the Mavericks showed: No, he can't turn it on. A team can even attack Bryant's weakness as a defender to dominate games. He constantly lost his man off of screens, CONSTANTLY, he wasn't able to close out. And it didn't even matter which player he guarded. Even Stojakovic, a guy who is on his last legs, was able to get enough seperation from Bryant. And if a 2nd round series doesn't let Bryant turn it on, it might very well not there anymore.

When I evaluate players I don't pay any attention to the All-Defense awards. For two reasons: First, there is no explicite All-Offense award. The other thing is that I always find cases in which I completely disagree with the choices. Last season the coaches gave 4 1st team votes to Jason Kidd. That was absurd, because the opponents teams exploited his weaknesses to guard a quick perimeter player. In this season I find Joakim Noah on the 2nd team, a guy who missed a lot of games and isn't such a good defender. Noah, as many high energy guys, gets overrated on defense. Hustle and effort alone doesn't make a good defender. On the Bulls I see 3 players being better defender than Noah. That would be Luol Deng, Ronnie Brewer and Omer Asik. It is a shame that a voting panel of coaches can't even get that right for a specific team. Noah and Rose got more votes than Luol Deng. That is an amazing thing here.
And that is just by looking at a specific team. If we look further we can see that Andrew Bogut got just 1 vote, one vote while Noah got 12 votes. So, Bogut missed some games, but Noah missed MORE games, Noah played LESS minutes. And when Bogut was on the court the Bucks had with 101.0 a better DRtg than the Bulls when Noah was on the court (102.8). But when those coaches can't even get the ranking on a team right, how are they able to decide between players from different teams?

The whole thing shows that we shouldn't base our opinion about the defensive strength of a player on the All-Defense team honors.

Let's put it this way, with just about every other accolade in sports, the voters seem to get it pretty accurate. Even if they are off at the top or there is a single outlier of sorts. (eg LaMarcus Aldridge didn't receive any MVP votes -- that's not reflective of his play or value at al. Rajon Rondo had 5 votes -- that's absolutely bonkers.)

With the all-defensive team, the results seem so disconnected from reality that it makes me wonder if something else is going here. Namely, that the lack of information (ie No defensive stats for decade after decade) renders the all-defensive team an exercise is whimsical reputations.

"Raja Bell slaps the floor - he comes to mind when I think of good defense."
"Joakim Noah screams a lot. I'll vote for him."

MVP and All-NBA teams receive more attention, logic and information. Joakim Noah missed 34 games. 34! Chicago's defense barely even suffered. This would be hugely relevant in even the most basic MVP argument, but apparently no one even thinks of it when submitting these ballots.

Conversely, Andrew Bynum received a single second-place vote. He played more -- so that can't be it. He had a higher DRB% -- so that can't be it. He blocks significantly more shots -- that can't be it. His team's DRtg was better when he was in the game (101.6 to 102.8) -- that can't be it. That's not even delving into more complex stuff. So...it's the hair??

When I look at the pool of players receiving all-d votes every year, 99% of the league's good defenders of there. The order is just whacko and certain reputation players who shouldn't even really be on there -- eg Kobe and Shawn Marion -- are disproportionately represented. The reality is, most people who watch basketball keep an eye on the ball. They have a huge internal databank of shots and passes, but not rotations, errors, blow bys, stops, saves, recoveries, switches, etc.

My feelings have been for a long time that if the stats aren't good, it's amazing how quickly the "guys who really know the game" get exposed as working primarily off of reputation, with the creation of the reputation having a ton to do with luck.

The NFL Pro Bowl is filled with stuff like that. Gary Payton was still considered the best perimeter defender in the game until it became clear that he was actually below average - the experts completely missed the fall that didn't happen over night.

We've also seen in basketball the experts rely on bad stats when good stats that aren't available. Larry Hughes leads the league in steals? Boom, All-Defense.

Criticisms about over reliance on stats often have a legit beef, but it just seems clear to me that where good stats don't exist, no human being seems to be able to have a good handle on players they don't see night in and night out - only a select few get that handle even with constant exposure.

ElGee wrote:With the all-defensive team, the results seem so disconnected from reality that it makes me wonder if something else is going here. Namely, that the lack of information (ie No defensive stats for decade after decade) renders the all-defensive team an exercise is whimsical reputations.

Is it really so disconnected from reality? At least, in ezPM, Bryant is the third highest rated guard:

It gets worse. This year, Javale McGee got votes for All-Defense. Yes, McGee blocks lots of shots, but he was actually a bad defender this season. Horrible rotations. Horrible instincts. Unable to hold position against anyone with strength. So excited about blocking shots that he probably led the league in goaltending calls.

I think your analysis is pretty bad. Here's a few balancing points to consider.

-SG might be the weakest defensive position. There are no SGs that stick out defensively as great.
-Kobe's net production allowed has ranged from good to elite over his career.
-Fisher is horrible by any stat I can see. Constantly having to make up for Fisher could be an explanation for Kobe's relatively poor +/- on defense.
-Adj +/- has a lot of problems, especially with the Lakers as I have shown.
-There are no other reliable defensive metrics you can base your opinion on because the defensive stats are generally terrible.
-Kobe did play a little worse this year. But in previous years the Lakers had a great defensive team, especially in the playoffs.
-When Kobe played with Farmar instead of Fisher the Lakers were good to extremely good defensively. Farmar is underrated but he isn't that great.
For example last year the Fisher/Kobe/Artest combination allowed 99.9 ppg in the playoffs while Farmar/Kobe/Artest averaged 85.6 ppg. This is exactly the result you would expect if Kobe was a great defender being dragged down by Fisher.

Maybe Kobe didn't deserve it this year. Even this year though his net production allowed is elite.
But every other year you can make a pretty good case that he did in fact deserve it.

laika wrote:I think your analysis is pretty bad. Here's a few balancing points to consider.

-SG might be the weakest defensive position. There are no SGs that stick out defensively as great.
-Kobe's net production allowed has ranged from good to elite over his career.
-Fisher is horrible by any stat I can see. Constantly having to make up for Fisher could be an explanation for Kobe's relatively poor +/- on defense.
-Adj +/- has a lot of problems, especially with the Lakers as I have shown.
-There are no other reliable defensive metrics you can base your opinion on because the defensive stats are generally terrible.
-Kobe did play a little worse this year. But in previous years the Lakers had a great defensive team, especially in the playoffs.
-When Kobe played with Farmar instead of Fisher the Lakers were good to extremely good defensively. Farmar is underrated but he isn't that great.
For example last year the Fisher/Kobe/Artest combination allowed 99.9 ppg in the playoffs while Farmar/Kobe/Artest averaged 85.6 ppg. This is exactly the result you would expect if Kobe was a great defender being dragged down by Fisher.

Maybe Kobe didn't deserve it this year. Even this year though his net production allowed is elite.
But every other year you can make a pretty good case that he did in fact deserve it.

-It's true that the list of great defenders is dominated by big men, and that that is relevant in considering just how ridiculous a vote for Kobe is.
-Also true that according to 82games, Kobe's opponent's production is typically below average (which is a good thing)
-Whatever issues that exist with Kobe's team, the fact is that APM tries to adjust for that and that Kobe never looks great by that metric.

In the end, the fact that my post is based on metric is a pretty legit knock on it so I feel you there.

Why don't I pay more mind to 82games's opponent stats? If we were just talking about a short time period, I wouldn't feel comfortable using any one stat. I do feel comfortable using APM to make statements that are pretty obvious. The 8 year numbers tell us Garnett is consistently unreal and Kobe is consistently average (which granted is probably slightly above average for a SG). Add in that there are plenty of accounts (including my eyes) that tell me that Kobe's not giving his focus on defense all the time, and it makes plenty of sense to me.

In some ways my article is a pretty rough thing to write. After all, if Kobe has a finite amount of energy, what's wrong with conserving it some? Where I become militant though is that I know how people will use those accolades in the future. And the reality is thought there isn't anyone who knows what their talking about who wouldn't tell you that the swingman who takes the tougher role for the Lakers is Ron Artest, and yet Artest has only 2 1st team defensive awards instead of 9 like Kobe. Positions are a factor here, but the defense of guards isn't so weak that anyone should be thinking that that the guy doing the most defensively as a 2 guard is his team's #2 defender of swing men.

Nivek wrote:It gets worse. This year, Javale McGee got votes for All-Defense. Yes, McGee blocks lots of shots, but he was actually a bad defender this season. Horrible rotations. Horrible instincts. Unable to hold position against anyone with strength. So excited about blocking shots that he probably led the league in goaltending calls.

I'd be interested to hear you give your feelings about McGee going forward. Do you see him as a building block for your team's future success? Do you think he's close to being a great defender if he just figures a couple things out?

Doctor MJ wrote:-It's true that the list of great defenders is dominated by big men, and that that is relevant in considering just how ridiculous a vote for Kobe is.
-Also true that according to 82games, Kobe's opponent's production is typically below average (which is a good thing)
-Whatever issues that exist with Kobe's team, the fact is that APM tries to adjust for that and that Kobe never looks great by that metric.

The problem is that some really great guard defenders are snubbed year in and year out. Where is Manu Ginobili? The guy NEVER won a defensive team award. That's amazing. In this season he played like an All-Defense 1st teamer, but was again ignored.

Ginobili's 6 yr defensive APM by Iliardi: +3.24
6 yr by Engelmann: +1.5
4 yr by Engelmann: +1.9
This season by Engelmann: +2.2

Ginobili since 2004/05 got overall 24 points in the voting, 4 times 1st team votes and 16 2nd team votes. And over that time span the numbers are showing that Ginobili was a better defender than Bryant. His opponent PER in average from 2005/06 to 2010/11 is 13.0, while Bryant's is 13.5. And the Spurs didn't hide Ginobili on the weakest opponents player like the Lakers did in this season. Bryant defended Keith Bogans against the Bulls. The Lakers during the Mavericks series decided to let Bryant defend Stevenson, Kidd or Stojakovic, because whenever he had to defend Terry Bryant was lost on defense. And I mean LOST. Watch him in the first half of game 4 standing around and defending NOBODY, because he completely lost his man. And it didn't matter whether it was Kidd, Terry, Stevenson or Stojakovic. In one sequence Bryant was on the complete opposite side as Terry and Stojakovic, he didn't even try to fight through the screens. Bryant showed a lack of fundamentals on defense. And that didn't start in this season. His numbers aren't any good, because he started to play that way back in 2003/04. And Phil Jackson even called him out for this. Those +/- numbers aren't surprising at all, if you would have followed his development on the defensive end closely. Yeah, Bryant was able to lock someone down, when he wanted, but in terms of being a good team defender Bryant wasn't there with the best. Bryant wasn't playing like a great overall defender for all those years, but still racked up 7! All-Defense 1st team awards. During the same time span a player who played fundamentally better defense, who was an overall better defender didn't even got on the 2nd team once. Ginobili's 24 points in the voting would have been enough to make 1 All-Defense 2nd team in any of those years, and he got that amount over 7 years. Incredible.

In this season we have Ronnie Brewer, Tony Allen, Andre Iguodala and Manu Ginobili playing better defensively than Kobe Bryant.

Name DRAPM PPP (rank)

Tony Allen: +2.1 0.77 (22nd)
Manu Ginobili: +2.2 0.81 (58th)
Ronnie Brewer: +2.9 0.80 (49th)
Andre Iguodala: +2.1 0.81 (58th)

Compare that to the -0.8 and 0.89 (212th) for Bryant. Bryant can't keep up with the points per play derived by video analyis from Synergy Sports and he can't keep up with the DRAPM values. We have 4 shooting guards being better defensively and playing like that as Bryant, only two of them got the 2nd team award while all 4 would have deserved the 1st team over Bryant EASILY.

Kobe Bryant's Net production in this season is misleading. He was matched up with weaker offensive players while Artest defended the stronger SG. No play-by-play will tell you that. 82games.com will tell you that Artest had the worse opp PER with 14.4 while Synergy Sports actually captured the different matchups and has Artest with 0.8 as the 49th best, better than Bryant. Artest had +1.4 by Engelmann in this season as his DRAPM.

And what will be the result in the future? People will claim that Bryant was a better defender than Artest or Ginobili even in 2011, because he got that All-Defense 1st team award and the other two nothing.

Doctor MJ wrote:

Nivek wrote:It gets worse. This year, Javale McGee got votes for All-Defense. Yes, McGee blocks lots of shots, but he was actually a bad defender this season. Horrible rotations. Horrible instincts. Unable to hold position against anyone with strength. So excited about blocking shots that he probably led the league in goaltending calls.

I'd be interested to hear you give your feelings about McGee going forward. Do you see him as a building block for your team's future success? Do you think he's close to being a great defender if he just figures a couple things out?

McGee is a monumental question mark. He has the athleticism to be a Dwight Howard-like figure in the league. But, he's immature, doesn't seem to have the drive to become great, and he also suffers from asthma. The big thing with him on the defensive end is a general lack of awareness/low hoops IQ. I think it'll get better with experience -- especially if he grows the willingness to listen to coaches. He made some progress late in the season. His rotations weren't a lot better, but he stayed down a bit more and didn't try to block EVERYTHING.

On the offensive end, the word that comes to mind quickest for me is "primitive." Film him in black & white and put him in those old-timey short-shorts and he'd look like someone from the early 60s. He did show some progress toward the end of the season, though. He has a nice swooping right-handed hook that's unblockable, and he showed some semblance of a drop step too. He likes to think he has a jump shot, but his results are horrific (and he's a lousy free throw shooter).

His athletic gifts (running and jumping) are so overwhelming that he can get on Sports Center nightly. But, he still has to learn how to play. During national team workouts last summer, Lamar Odom (when asked about McGee) said McGee needs to learn that the game is "basketball, not run and jump."

Good thing about McGee is that he plays hard most of the time. He's in the weight room trying to get stronger. Don't know how much on-court work he's doing (he needs a lot). His effort isn't lacking, it's just that sometimes his efforts are misguided, and he has a stubborn streak that makes it difficult for coaches to get through to him at times.

My hope is that his experience last season has him working harder on his game and that he'll become an All-Star and building block. Even if that doesn't happen, he's good enough right now to be an 18-20 minute reserve big. That would be a colossal waste of that athleticism and body, though.

I feel like every NBA award has lost meaning...

If Greg Monroe had McGee's physical gifts, he'd be a super-star.

EvanZ wrote:

ElGee wrote:With the all-defensive team, the results seem so disconnected from reality that it makes me wonder if something else is going here. Namely, that the lack of information (ie No defensive stats for decade after decade) renders the all-defensive team an exercise is whimsical reputations.

Is it really so disconnected from reality? At least, in ezPM, Bryant is the third highest rated guard:

Code: Select all

Dwight Howard   5.08
Tony Allen   4.14
Andrew Bogut   3.40
LeBron James   3.33
Tim Duncan   3.21
Rudy Gay   3.14
Ben Wallace   3.13
Andrei Kirilenko   3.08
Andre Iguodala   3.03
Carlos Delfino   2.96
Kurt Thomas   2.93
Ron Artest   2.90
Elton Brand   2.68
Paul Pierce   2.66
Luol Deng   2.62
Kobe Bryant   2.52
Kevin Garnett   2.51
Marc Gasol   2.50
Emeka Okafor   2.40
Josh Smith   2.32

But aren't you using counterpart data? That means other players, like Artest, could be guarding those players. It also means, more importantly, that the wing players with great bigs behind them will benefit more.

Kobe Bryant gets blown by constantly. The Lakers give up the fewest points in the league in my database after a defensive error, because they usually have 15 feet of wingspan waiting in the paint. That's going to be reflected in those numbers.

And it is disconnected from reality. Bryant, by design, rests on defense. He guards the weaker wing usually. The whole idea that people who know the NBA IGNORE this, by definition, is a disconnect from reality. If the argument is "Bryant COULD be the best wing defender if he wanted to" -- something there is no evidence for -- it's still nonsensical to reward that. It implies Michael Jordan should have been on the all-nba 1st team defense in 1994.

My 12-month sample had plenty of Bryant defense. Sample size and error are not an issue. He comes out just above average, behind wings Barnes, Pietrus, Grant Hil, Fields, Durant, Pierce, Bogans, Wade, Ginobili, Salmons, Tony Allen, LeBron. Strictly at the 2, Wade, Ginobili and Tony Allen are CLEARLY better defenders. I don't know any plausible argument I could genuinely muster for Kobe there.

As for the rest of the league at shooting guard, Ronnie freaking Brewer was light years ahead this year. Is Bryant a better defender than Thabo Sefolosha? What about Paul George?

I mean, even if you restrict the all-defense selection to *only* the 30 starting SG spots, and even if you muster up an argument that Kobe is in top-15 (or even top-10), he's still miles away from deserving a spot on that team. He wasn't even on my radar...it's about as realistic as voting Keith Bogans for DPOY.

ElGee wrote:
My 12-month sample had plenty of Bryant defense. Sample size and error are not an issue. He comes out just above average, behind wings Barnes, Pietrus, Grant Hil, Fields, Durant, Pierce, Bogans, Wade, Ginobili, Salmons, Tony Allen, LeBron. Strictly at the 2, Wade, Ginobili and Tony Allen are CLEARLY better defenders. I don't know any plausible argument I could genuinely muster for Kobe there.

What's interesting is that the list is created by NBA coaches, not media. Kobe received 13 first-team votes, the fewest of any on the first team. 10 coaches didn't vote for him for 1st or 2nd team. Interestingly, the first three who missed making the list are Rose, Wade, and Westbrook. Ginobili only received 2 votes. Are we perhaps missing something here? Could 28 coaches be completely wrong about Ginobili, but ElGee has it right? It looks like 9 out of the 10 picks are reasonable to everyone, but apparently not Kobe.

What exactly is the motivation for coaches to pick Kobe, but not Ginobili? These things are just curious to me. The thing is that the coaches probably have a lot of the same data - and even more sophisticated data - than what ElGee has. They do this for a living. They create game plans to go against certain players (or not). Maybe Kobe brings other things that we can't see, but the coaches do? I wish I knew. Maybe it's simply that the coaches think this is not about one particular year, but the entire body of work over a career - but that doesn't make sense in the case of Allen. I can't honestly think of a good reason that they wouldn't genuinely pick the player they thought was the best defender.

Ok, here's one shot. Maybe the coaches who voted for Kobe remember him having particularly good games against their team. Maybe he even had just one spectacular defensive effort, and that was enough in those coaches minds to give him their vote. This would just be another form of confirmation bias. They know Kobe has a reputation of being good on defense (they have a scouting report that tells them so, I assume), and when he shuts down one of their stars, that's what they remember. Too bad they don't make the votes public, so we could check this hypothesis.

I can't tell if you're being condescending.

Please don't act like coaches who are slipped an all-d ballot put the effort into that they do their teams winning. Nor should we assume that the coaches voting for this award are designated to defensive scouting/analysis. I've been in NBA locker rooms and read through scouting packets -- I've never seen anything suggesting that teams are hyper-aware of the defensive PERFORMANCE of opposing individuals. And even if they were, if they took the time to parse them out properly when handed a ballot with hundreds of names to choose from.

Think about that: Scouting reports are usually about opposing offense and team tendencies. They may say stuff like "Miami bigs like to show hard on the PnR." They speak to tendencies -- "Bibby will try and front you in the post." And with Kobe, I imagine, the note that he goes for steals, jumps lanes well and the Lakers funnel you to their bigs. (Boston discovered Wade trails behind the play on missed drives to the hoop.) I don't see how we can be sure the people voting on this are taking it seriously or are the ones who have used more complicated information to make more informed macroscopic decisions...when it has nothing to do with their job to make such decisions.

Also, the numbers I'm citing are numbers, they aren't *me.* I'm just looking at what happens over the course of the games, in the same way one would tally FG% and points. What does a totally non-causal measurement, like APM, say about this?

For me, as I said, I don't know how to watch all these players play and muster up a cogent argument for Bryant over any of them. I use all data I can, some of which is totally outside the scope of available statistics. And I hear all kinds of people who DO watch defensive film and are defensively minded at that level (eg Jeff Van Gundy) constantly comment on how overrated Bryant's defensive impact and how good of a defender Ginobili is.

I can come up with plenty of other theories, too, that have to do with, reputation, biases, exposure, nationality, etc. For goodness sakes, look at some of the votes (Bynum 1 vote, Joe Johnson 2 and a 1st place, Ron Artest a 1st place vote...which means there's someone on that magic voting panel that thinks Artest is a higher impact defender than Kevin Garnett or LeBron James.)

ElGee wrote:I can't tell if you're being condescending.

Not at all. Sorry if I came off that way.

ElGee wrote:
Also, the numbers I'm citing are numbers, they aren't *me.* I'm just looking at what happens over the course of the games, in the same way one would tally FG% and points. What does a totally non-causal measurement, like APM, say about this?

Some of your stats are subjective. Specifically, defensive errors. All humans have subconscious (and often conscious) biases. A proper scientific experiment would be to blind yourself from the player, so you wouldn't know who you were tracking. Obviously, that is not feasible. Still, when you tally "blow-bys" and "missed rotations", that's subject to error. I can't imagine that if we got 5 guys in a room and watched a game that there would be universal agreement without having some discussion first. If you are claiming otherwise, I would be skeptical. If a Laker fan were to tally these stats, we might see different results than if a Miami or San Antonio fan did.

This is not to say I believe you are intentionally attributing more or less errors to any particular player. But it is very easy for humans to let even the most subtle biases influence the outcomes of their measurements.

I happen to be very acutely aware of these issues, as someone who is a scientist in academia doing research and supervising students. I make it very clear to them the issues with subjective vs. objective measurements and how we must fortify ourselves against these kinds of biases. It's not an insignificant problem.