Bret Lagree's brilliant blog on Joe Johnson.

[mrhonline]

Link

Here's the summary, but read the whole thing:

Joe Johnson had a bad defensive season. He wasn't used optimally but the combination of that with his lack of a measurably positive defensive impact in 2008-09 as well, his below average offensive rebound rate, his sharp decline in both free throw and block rate, the paucity of defensive plays made, an increasing inability either to get to the rim or create space for a mid-range jumper, and the volume of minutes he's played calls his future into serious question. Every athletic marker in his statistical profile draws a red flag.

These are the kind of numbers that indicate a player is unlikely to translate his collegiate production to the NBA. However one feels about the validity of Johnson's current reputation, it's difficult to see how he will maintain his established level of production into his mid-30s. Buyer beware.

[niremetal]

Link

Here's the summary, but read the whole thing:

You're so predictable that it's sad. That post is already up on the imports forum, but you want everyone on HS to see that someone else agrees with you on JJ, so you post it on the front page and call it "brilliant." Bravo.

Edited May 21, 2010 by niremetal

[mrhonline]

A substantive rebuttal would've been a more mature response, niremetal.

[niremetal]

A substantive rebuttal would've been a more mature response, niremetal.

Why should posts without any substance deserve responses with substance? All you did was re-post something already up on the imports page and call it "brilliant."

Edited May 21, 2010 by niremetal

[lethalweapon3]

I disagree with the premise of defensive impact lacking from Joe... or perhaps I agree, in the sense that his defensive impact is immeasurable.

He has to create his own shots on one end of the floor, then get back and guard not only his own man but the opponent's hot PG/SF on the other end, largely due to Bibby's ineffectiveness and Marvin's inconsistency. Rio couldn't be on the floor for long due to a lack of offensive aptitude, so Joe logs heavy minutes that reduce his "efficiencies."

It shows up in the playoffs where our star player has to chase down the likes of Brandon, Jameer, LeBron, D-Wade, often on his own, while opposing teams can sic other players (multiple players during iso's) on him at the opposite end. He's been compensating for our defensive shortcomings pretty much since he got here, to the detriment of his per-minute stats.

I do agree Joe does need to get to the line more, though. Probably another 5-6 points the Hawks are leaving on the table every night.

~lw3

[mrhonline]

Why should posts without any substance deserve responses with substance? All you did was re-post something already up on the imports page and call it "brilliant."

Nice deflection. Still waiting on your rebuttal to what LaGree said.

(IOW, your personal attacks won't work on me).

[Brotha2ThaNite]

I'm not to call JJ Great.... but I think JJ was used in the wrong way. He had the ball in his hands too much. If you put a PG on the court (LETTING THAT PG BRING THE BALL UP THE COURT AND RUN THE OFFENSE) like maybe teauge you will see a totally different JJ.

I hated to see JJ bring the ball up the court. Woody did this too much. I remember even when we were getting killed against the Magic. He put Teague in but still ran the offense through Crawford. Crawford brought the ball up court. Us the PG for what a PG does best. Let JJ play the Reggie Miller role..... He will kill.

By the way, JJ avg 20, 5 and 5.... it was like only 5 ppl in the league to do that, right? So, how is his stats bad? :cant believe:

[Buzzard]

I disagree with the premise of defensive impact lacking from Joe... or perhaps I agree, in the sense that his defensive impact is immeasurable.

He has to create his own shots on one end of the floor, then get back and guard not only his own man but the opponent's hot PG/SF on the other end, largely due to Bibby's ineffectiveness and Marvin's inconsistency. Rio couldn't be on the floor for long due to a lack of offensive aptitude, so Joe logs heavy minutes that reduce his "efficiencies."

It shows up in the playoffs where our star player has to chase down the likes of Brandon, Jameer, LeBron, D-Wade, often on his own, while opposing teams can sic other players (multiple players during iso's) on him at the opposite end. He's been compensating for our defensive shortcomings pretty much since he got here, to the detriment of his per-minute stats.

I do agree Joe does need to get to the line more, though. Probably another 5-6 points the Hawks are leaving on the table every night.

~lw3

Not to mention potential foul problems for the player trying to guard him and the big playing the post ( Howard anyone?)...this has always been my biggest gripe with JJ; but at the sametime, it is also a big part of the reason IMO he has never come close to missing as many games as a player like Wade despite playing equally massive minutes....

Edited May 21, 2010 by Buzzard

[niremetal]

Nice deflection. Still waiting on your rebuttal to what LaGree said.

(IOW, your personal attacks won't work on me).

And I'm still waiting for your response on the last thread where you demonstrated your M.O. of "let me start with my conclusion that we shouldn't re-sign JJ and search for 'evidence' that supports it."

I've talked about this plenty. Here is the Cliff Notes version in case you forgot. JJ is an all-NBA player and, in my view, would be even better offensively in a system that tried to create some open looks for him instead of demanding that he create all his shots himself. His game is likely to age well because he relies very little on his speed/quickness and is an excellent catch/shoot and off-screen scorer when he is given the chance to do those things. His defense has been adversely impacted by the fact that he starts off guarding the opposing team's PG and often ends up guarding the opposing PF or C off switches. The on/off data for points allowed is an idiotic measure because it doesn't account for the relative offensive quality of opposing players or the relative defensive quality of the players who come in and replace him (which is why Bibby and Josh had identical On/Off defensive differentials). In any case, it's hard to find people who aren't haters and who don't drool over Hollinger-esque "data" who says that JJ isn't a very good defender.

Edited May 21, 2010 by niremetal

[Guest]

The on/off data for points allowed is an idiotic measure because it doesn't account for the relative offensive quality of opposing players or the relative defensive quality of the players who come in and replace him (which is why Bibby and Josh had identical On/Off defensive differentials).

In theory it is controlling for opposing quality because we should have a large amount of substitutions and on average over the 48 minutes of 82 games played each player should be facing the same opposing quality. And what its goal is to do is to rate the relative defensive/offensive quality of the players that the plus minus is measuring. To say it doesn't control for relative defensive quality of the players who replace is a little foolish, +/- is trying to measure that so of course it isn't controlling for that. The only way to measure it is for there to be changes and then record the offensive or defensive ratings after these changes.

The concept is not the problem. The concept works and is used in statistics and regressions quite a bit. No one in statistics would question the method. However, the problem is that in statistics there is the assumption of exogeneity. Plus minus will give you the results you want if substitutions are exogenous. But the coach controls substitutions so we don't get exogenous substitutions and our results are biased. So the simple +/- is flawed, although I have read in many places that there are people working on this problem and creating more advanced +/- metrics to correct it.

[niremetal]

In theory it is controlling for opposing quality because we should have a large amount of substitutions and on average over the 48 minutes of 82 games played each player should be facing the same opposing quality. And what its goal is to do is to rate the relative defensive/offensive quality of the players that the plus minus is measuring. To say it doesn't control for relative defensive quality of the players who replace is a little foolish, +/- is trying to measure that so of course it isn't controlling for that. The only way to measure it is for there to be changes and then record the offensive or defensive ratings after these changes.

The concept is not the problem. The concept works and is used in statistics and regressions quite a bit. No one in statistics would question the method. However, the problem is that in statistics there is the assumption of exogeneity. Plus minus will give you the results you want if substitutions are exogenous. But the coach controls substitutions so we don't get exogenous substitutions and our results are biased. So the simple +/- is flawed, although I have read in many places that there are people working on this problem and creating more advanced +/- metrics to correct it.

My point was unclear due to bad sentence structure. I wasn't trying to imply that the metric doesn't account for the quality of the sub and the quality of opposing players (individually). What I should have said was that it does not control for the relative quality of their subs as the subs are used and taking into account who the subs are asked to guard, which is (in part) a function of "the relative offensive quality of opposing players." A sub whose minutes are reduced or his substitution pattern is altered in games where the player he would be defending is a good scorer would have a deceptively high simple +/-.

And yeah, there are supposedly a bunch of models out there that teams hold essentially as trade secrets that supposedly do regressions that account for the exact matchups the players faced on the floor. The biggest problem, in my view, is that no statistical regression can capture the impact of simply how the player is being used in the system (or lack thereof) in which he operates. We can look at statistics that tell us how JJ did in 2009-2010 under Woody. We cannot project how he would do playing in a motion offense or an offense that worked him off the ball, in a non-switching defense, or when he plays alongside a better defensive PG.

No stat can project how JJ would perform playing alongside Derrick Rose or even under D'Antoni without Nash/Marion/Amare alongside him. No regression can project how Marvin or Josh would do in systems that defined their offensive roles more clearly. But to me, that's what makes basketball so great. It's the most difficult sport in the world to quantify.

Edited May 21, 2010 by niremetal

[Guest]

No stat can project how JJ would perform playing alongside Derrick Rose or even under D'Antoni without Nash/Marion/Amare alongside him. No regression can project how Marvin or Josh would do in systems that defined their offensive roles more clearly. But to me, that's what makes basketball so great. It's the most difficult sport in the world to quantify.

Right, there isn't a statistic that can measure how a team interacts with a new coach unless you have a coaching change. But can't we look at a player's performance over the tenure of a coach and be able to gauge an incline/decline in his production?

I disagree with you about basketball being the most difficult sport to quantify. Basketball has so many measurable statistics that we can approximate different things. If you look at Soccer, you can see ball control, fouls, shots, saves, and goals. Very few statistics which makes it hard to quantify. Football is the same way except for maybe the RB and QB positions. Hockey really doesn't have as many measurables and if you look at the +/- for Hockey you run into the same endogeneity bias that you do in basketball.

Basketball certainly lends itself to statistics. Its a harder thing to quantify than baseball but out of all the sports that people follow in general, it is actually one of the easier to quantify.

[niremetal]

Right, there isn't a statistic that can measure how a team interacts with a new coach unless you have a coaching change. But can't we look at a player's performance over the tenure of a coach and be able to gauge an incline/decline in his production?

I disagree with you about basketball being the most difficult sport to quantify. Basketball has so many measurable statistics that we can approximate different things. If you look at Soccer, you can see ball control, fouls, shots, saves, and goals. Very few statistics which makes it hard to quantify. Football is the same way except for maybe the RB and QB positions. Hockey really doesn't have as many measurables and if you look at the +/- for Hockey you run into the same endogeneity bias that you do in basketball.

Basketball certainly lends itself to statistics. Its a harder thing to quantify than baseball but out of all the sports that people follow in general, it is actually one of the easier to quantify.

Again, hyperbole. It's the most difficult sport that I give a damn about to quantify :snowballfight:

Then again, there isn't much competition on that front. The only sports I follow are basketball, tennis, football, and baseball (in that order). I've had many debates with fellow geeks on whether basketball or football is easier to quantify. Suffice it to say that I think it's close, but I think basketball is more difficult for rather impressionistic reasons.

[Guest]

Then again, there isn't much competition on that front. The only sports I follow are basketball, tennis, football, and baseball (in that order). I've had many debates with fellow geeks on whether basketball or football is easier to quantify. Suffice it to say that I think it's close, but I think basketball is more difficult for rather impressionistic reasons.

This will be very off-topic, but I think tennis is an unbelievably interesting sport to quantify. One thing I have noticed are that people would claim that Tennis is a "mind-game" and that your mental focus is so necessary to winning that you cannot quantify it. Well I don't disagree that Tennis is one of the most taxing sports on your mind, but that doesn't mean you cannot quantify it. It leads itself to be observed in a time-series manner, where what happens in the past predicts what will happen in the future. So as Andy Roddick loses one point its has an effect on his next point but it isn't very large. But if he loses two points in a row, it has a bigger impact. And so on and so forth where you can observe a snowball effect. It isn't necessarily Andy losing the 1st point that ends up in a meltdown, or the game losing point but each point has a certain impact. That is something you can measure, although it gets tricky with distinguishing between Sets and Games and if you want to allow Games to carry over impact and so on.

[NJHAWK]

Link

Here's the summary, but read the whole thing:

While Joe seems to always have some kind of nagging injury that he plays through he has never had a serious injury. I think eveyone knows that by the 5th or 6th year of a potential deal that JJ will most likely be a bad contract. The thing is the last year of a bad contract is almost always a trade asset :beer11:

It is a scary read I'll admit that but not as scarry as JJ just walking and leaving this team with little or no compansation.

[Diesel]

Link

Here's the summary, but read the whole thing:

I read this and it's basically an agenda against Joe.

So let me speak for why we should resign Joe...

1. None of those mighty stats tell the whole story.... THIS WAS NOT DONE IN A CLOSET.

Statheads like to point out stats and forget that Joe Johnson was the focus of every defense we played. Nobody is trying to stop Josh, Horf, or Bibby. They all want to stop Joe and Woody would give them Joe in Iso form. With all that pressure, Joe still is a 4 time allstar. If Joe ever gets with a real PG and a good coach.... it's lights out.

2. We lose Joe for Nothing, we have NOTHING.

Right now... regardless of how embarrassing we were, we are one of the top 8 teams in the league. Personally, I would say 5th best. We lose Joe for nothing. That's definitely gone. We have said at length how there's no more money to spend on big free agents if Joe Leaves. We'd be stuck with our mistakes (Bibby, Marvin)... and have to learn to wait another 5 years before we can climb out of mediocrity. Why would we do that? Why would we put ourselves in that hamstring?

3. Joe was our best defender.

Bring your stats up if you want. All I know is that when we needed somebody shut down.. Joe did it. In the Milwaukee Series, you saw Joe Shut down Jennings and Salmons at different points. Call him unathletic, call him slow, call him whatever you want. He was still our best.

Lose Joe for nothing and we lose a tradable marketable piece.

Star players want to play with other stars. Joe is our star. I would love to see us get a great coach (AJ).. Trade Smoove/Marvin for Bosh SNT. Then rip up the scene with Joe Bosh and Horf... No Joe, we lose that chance. Plus, if we get Joe signed, we can trade Joe. The teams that really like Joe now, will still like him in Feb.

[Sothron]

I agree with Diesel. We can't just let Joe go for nothing. We don't have the cap room to bring in anyone close to his production. I'm not opposed to keeping him or trading him but the absolute worst thing we could do is lose him for nothing.

[99PROBL3MS]

You can certainly allude to a trend of better/worse production based on 5 or 6 years in the same system. I just think that JJ would be much more effiecient in a different offense. Put him in a Flex like Utah's, a triangle like L.A.s , a motion offense. Heck anything but this iso (one-on- five Barkley likes to call it) and his production could sky rocket. As nm said, the fact that his game isn't based on athleticism makes me think his game will age well as long as he is in a system besides the iso. He has never been much of a stat stuffer defensively but his overall defensive IQ is good and his anticipation is top notch. Neither of those can be measured statistically.

[Guest]

1. None of those mighty stats tell the whole story.... THIS WAS NOT DONE IN A CLOSET.

Statheads like to point out stats and forget that Joe Johnson was the focus of every defense we played. Nobody is trying to stop Josh, Horf, or Bibby. They all want to stop Joe and Woody would give them Joe in Iso form. With all that pressure, Joe still is a 4 time allstar. If Joe ever gets with a real PG and a good coach.... it's lights out.

Joe has continually been the focus of opposing defenses every year he has been here. Its not like all of the sudden he is being double-teamed now. The "statheads" do account for this. If how opposing defenses approach Joe had changed then you might have a point as to why we see declining stats. But we haven't seen that, so you are attacking something that the "statheads" don't dispute. Joe being the focus of other defenses is accounted for because all statistics are based off of how Joe performs given this team.

If Joe had declining stats and one year he was not double-teamed then the next year he was double-teamed then you could make a case that its the double-team that caused the decline. But Joe may have actually been double-teamed less this year than last because Horford improved offensively and Jamal was an upgrade over Flip. If anything, double-teams went down for Joe if not stayed the same. Yet we see declining statistics.

[TheNorthCydeRises]

Bret's thinking is flawed in so many areas. He cites that his defense is down, but he also knows ( and says ) that JJ was asked to guard PGs, as well as SGs and SFs. Even a guy like Kobe isn't asked to guard PGs on a nightly basis ( although he does step up to the plate and do it sometimes )

JJ is getting paid moreso off of what he's worth NOW, than what he's going to be worth 5 or 6 years from now. If we have a "window of opportunity" to win anything, it's not going to be based off of what this team will look like 5 years from now, it's going to be based off of what this team looks like in the next 1 - 3 years.

People have to stop looking at the fact that JJ won't be worth 18 - 20 million when he goes into year 5 of his deal, as basis on not to re-sign him. At that time, his salary number could become an asset to a team either looking to add a veteran piece to a championship squad, or to a team looking to obtain major cap room.

At the very worst, if ATL gets a legitimate PG in here, you're looking at JJ playing a Vince Carter type role in year 4 or 5 in which he may not be the focus point of the offense anymore, but still is the #2 or #3 option on the squad that has the ability to be the #1 guy at any give time.

Bret correctly cited that JJ is taking less long jumpers and operating more in that 10 foot range ( with the floater ). Well, this also indicates that he may be transitioning his game to more of a SF role, and less of a SG role. I've always compared JJ to a slightly less talented version of Paul Pierce. To me, they play almost the exact same game, with the exception of Pierce being able to draw a lot more fouls. But they're both versatile players than can do a lot of things.

And I don't think it's a coincidence that with the addition of Rajon Rondo to take the ball handling duties away from Pierce, that Pierce over the last 3 years has become a more efficient shooter across the board, especially from 3 point range. He gets A LOT of spot-up looks from 3 these days.

All in all, we pretty much have to re-sign dude, or risk taking a step back. While Horford has improved, the jury is out BIG TIME if he can become a legit go-to scorer. The same goes for Josh Smith. Without a PG here, someone is still going to have to step up and be able to score off of their own created shot.

Not having JJ or a replacement go-to scorer is far worse than overpaying the guy.