The Phoenix Suns Defense

[SunsFan89]

45 games in what are your overall thoughts on the Phoenix Suns Defense so far this season?

 

[AzStevenCal]

45 games in what are your overall thoughts on the Phoenix Suns Defense so far this season?

We play decent defense until Green or Marcus take the court. Well, actually you have to put Wright in that group too but he's still trying to figure out how to fit in. Green and Marcus are among the absolute worst defenders in the league this year and I see no excuse for it. Bledsoe, Tucker, Len and Keef all defend fairly well despite the abuse number 11 gets here. When Green and Marcus are both on the court, we're probably the worst defensive team in the league.

Steve

 

[elindholm]

45 games in what are your overall thoughts on the Phoenix Suns Defense so far this season?

It's lousy. The Suns have some above-average individual defenders, but the team defense is poor because there's precious little group awareness or strategy. Offensively, the Suns rely mainly on one-on-one, and it's as though that the only kind of offense they expect the opponent to muster. Moving without the ball or a succession of quick passes will expose the Suns defensively almost every time, since they aren't accustomed to thinking about those possibilities on their own side of the ball. And you never see anything that suggests a plan like, "When so-and-so has the ball on the right low block, we're going to force him middle so that Guard X can drop down and double." It's every man for himself, mainly, and if you get beat, hope one of the guys gambling for a shot block comes through.

 

[sunsfan88]

There's just zero communication defensively. Nobody other than Tucker ever talks.

I read a recent article in which Derrick Rose ripped the Bulls players for effort and not communicating on defense and then they came out and blew out San Antonio and beat Dallas.

I think if Tucker called out our guys on communicating defensively, it could go a long way.

 

[Errntknght]

I'd call it a work in progress, mainly because of Len being on his learning curve. Also Wright is being fitted into the scheme. I don't think it destined to be terribly good - offense seems to be the deciding factor in who plays and who they pursue in the draft, trades and FA. (I'm counting rebounding as part of defense.) P.J. is somewhat counter to that but he is valued partially for his 3 pt shooting and he is quite efficient on the offensive end.

Even though I don't go on about it, I consider the offense orientation to be a serious blunder when building a team. Primarily because offense costs a lot more. Three PFs who can defend and rebound were given away this year - Ed Davis, Tarik Black and Jeff Adrien. Heck, Adren was waived by Minny and not picked up yet - I've been watching for him to surface. If their offensive and defensive capabilities were interchanged they'd have gotten decent contracts. Of course, you want to have a balanced team but you can get good much quicker if you start with defense. If you start with a focus on offense you rarely ever get around to defense... now and then you do stumble onto a Raja Bell or a Maurice Lucas, like once every 20 years.

 

[Phrazbit]

You have a team with a mostly poor defenders and you're going to have a poor defense. But its a young team and the roster is a work in progress. There is potential with a few guys to be elite at their positions defensively but we have quite a few that clearly butter their bread on the other side of the ball. I expect that over the next year we will a couple of those guys depart the club.

Especially Green. I think the defense has potential to improve markedly once he is gone. We could lose him tomorrow to the rapture and I think it would be addition by subtraction. His on court/off court splits are alarming. There is a game about every other week where he catches fire and gives us a big big push, but on the other end, in his role as a doormat, he is far more consistent.

 

[slinslin]

Gerald Green 104 0f 104 SGs, Goran Dragic 86 of 104, Brandan Wright 94 of 100 PFs and Marcus Morris 81 of 85 SFs are by advanced stats the worst or close to the worst defensive players in the league.

Brandan Wright has not really changed anything, in fact he has been pretty bad and is proving that he should not take minutes from Plumlee even.

On the other hand Markieff ranks 6th among PFs, Bledsoe 5th among PGs, Tucker 7th among SFs and Len 11th among 76 Cs.

Isaiah Thomas is middle of the pact.

 

[Son of RedRage]

We have a defense? I sure as hell never noticed.

 

[Errntknght]

Gerald Green 104 0f 104 SGs, Goran Dragic 86 of 104, Brandan Wright 94 of 100 PFs and Marcus Morris 81 of 85 SFs are by advanced stats the worst or close to the worst defensive players in the league.

Brandan Wright has not really changed anything, in fact he has been pretty bad and is proving that he should not take minutes from Plumlee even.

On the other hand Markieff ranks 6th among PFs, Bledsoe 5th among PGs, Tucker 7th among SFs and Len 11th among 76 Cs.

Isaiah Thomas is middle of the pact.

What site and stats give these kind of defensive rankings - or overall rankings, if thats what you are quoting??? Its not that I don't trust you slin, I would like to look them over.

 

[Phrazbit]

What site and stats give these kind of defensive rankings - or overall rankings, if thats what you are quoting??? Its not that I don't trust you slin, I would like to look them over.

I think he is using ESPN's RPM stats. Which I have mixed feelings about. Its a good stat for comparing within one roster but IMO a bad stat for comparing guys across the league. The stat is based almost entirely on weighting one guy against the guys who are his teams alternatives at the same position(s). If a team has particularly poor reserves in the role you see will extremes in the stat, the Suns numbers are a perfect example. On the other hand a guy like Tony Parker looks like a really sub par player in RPM, but its because the Spurs have such strong rotations behind him that it drags his off court rating way down.

 

[Errntknght]

Thanks Phrazbit, I'd found that site once a good while back but I haven't been able to find it again - almost didn't find it this time because its a tiny link off to the side. The rankings he gave match the DRPM rank, not the Wins Above Replacement, which I've never found very useful. From their footnotes RPM is plus-minus adjusted for who else is on the floor (own and opp team) plus 'other' factors. From what I've read about this, they use repeated statistical regressions to the do the adjustment which means that no one truly understands the stat. I've used regression analysis a few times in my work and I was always a little queasy about it... quite small changes in the data can have large effects - rather like the butterfly effect we read about. I'd rather live wilth the problems of raw +/- because at least I understand that.

I like the 82-game method of using 'production' against to measure defense... I'm not thrilled with PER like stats but they're understandable, or would be if they gave the formula they use. Green's 16.9 PER against is not going to put him at the bottom of the SG's. Tony Parker doesn't good this year on the +/- stats even if you ignore the OFF floor column - his per against is 17.8 and his own per 15.7.

I wish they had a table that listed the PER for and against for all players then I could copy them all and import them into my database.

 

[Phrazbit]

I think Green's PER against is misleading because he frequently is not actually covering the other team's SG, rather he is covering whoever the oppositions weaker wing player is, however, the stat merely reflects his PER against as that of his positional counterpart. So some of his not quite terrible appearance in that stat is the result of Bledsoe or Tucker covering his 'statistical' man.

One stat that does not lie about Green is his on court off court stats. The Suns are a disaster defensively when he enters the game. The Suns surrender more in rotations that feature Green than they do with any other player, and the only guy even close to him in terms of wrecking our team defense is Marcus Morris. A significant reduction in the role of those two players would likely improve our defense dramatically. Seeing as they obvioulsy drag the rating down of everyone one who plays with them, I'd like to see what our points against per 100 possessions are in lineups that feature neither one of them compared to lineups that feature both. I suspect its a difference of nearly 10 points per, maybe more

The Marcus thing is a tricky situation. It sucks that he is so strongly tied to Markieff, because Markieff is good... but I'm not sure he is good enough that it justifies giving 20+ minutes per game to his replacement level player of a brother, but I'm pretty sure Markieff would throw a tantrum if we benched his brother. I hope the team isn't married to them. I'd swap the two of them for Greg Monroe in a heartbeat... and I'm not a big Monroe fan.

 

[AzStevenCal]

I think Green's PER against is misleading because he frequently is not actually covering the other team's SG, rather he is covering whoever the oppositions weaker wing player is, however, the stat merely reflects his PER against as that of his positional counterpart. So some of his not quite terrible appearance in that stat is the result of Bledsoe or Tucker covering his 'statistical' man.

One stat that does not lie about Green is his on court off court stats. The Suns are a disaster defensively when he enters the game. The Suns surrender more in rotations that feature Green than they do with any other player, and the only guy even close to him in terms of wrecking our team defense is Marcus Morris. A significant reduction in the role of those two players would likely improve our defense dramatically. Seeing as they obvioulsy drag the rating down of everyone one who plays with them, I'd like to see what our points against per 100 possessions are in lineups that feature neither one of them compared to lineups that feature both. I suspect its a difference of nearly 10 points per, maybe more

The Marcus thing is a tricky situation. It sucks that he is so strongly tied to Markieff, because Markieff is good... but I'm not sure he is good enough that it justifies giving 20+ minutes per game to his replacement level player of a brother, but I'm pretty sure Markieff would throw a tantrum if we benched his brother. I hope the team isn't married to them. I'd swap the two of them for Greg Monroe in a heartbeat... and I'm not a big Monroe fan.

Monroe is playing much better now than in the past but I still think Markieff is at least his equal. Keef plays better defense (minus the rebounds) and is a much better clutch player. When you consider salary, Keef becomes the far more valuable asset. But even still, I might make that deal too.

I was expecting improvement from Marcus but with his complete lack of focus on defense along with his insistence on forcing shots and making his customary lazy pass ever game, I'm ready to say goodbye to him. I might be overreacting but he's just been so frustrating to watch this season. I really think that Marcus and Green together could turn Chicago into a below average defensive team.

Steve

 

[Phrazbit]

All things being equal I'd probably take Morris on his salary over Monroe on what he will be being paid next year... however, if having Markieff comes with the requirement of Marcus dragging us down with regular rotation minutes, then I'm over it.

Maybe Marcus will make a big leap, but outside of arguing with refs I don't see the same energy and aggression from Marcus.

 

[Errntknght]

I found something interesting in going over the Suns plusminus data on 82games.com.

One of the things you can look at there are the top 20 floor-time lineups. Those 20 lineups account for 1212 minutes of the teams approx 2182 or 55%. The teams net +- for the 45 games they have included (last update to the site is 1/23/15) is 117 but the net +- for those 20 lineups is 222 which means that all our other lineups have a net +- of -105. Not surprisingly when you calculate the individuals +- just on those 20 lineups everyone has a much higher +-/48 result, except for Markieff Morris - his is higher but only modestly so 9.8 vs 8.4.
Gerald Green on the other hand has +- of 14.4 vs 0.2, and Thomas has 16.8 vs 4.1 The reason is fairly clear - 83% of Markieff minutes occur within those 20 lineups while Gerald and Isaiah get only 37% & 32% of their minutes in those lineups.

There is a chicken and egg thing going on, of course - Markieff is a good part of the reason those lineups get floor time. Dragic and Bledsoe are more like Markieff than Green and Thomas with 69% & 67%, resp., of their minutes occuring in the top 20 lineups. Their +- rating don't change so dramatically either, 7.7 vs 3.4 for Goran and 8.8 vs 3.7 for Eric.

Remember these lineups were not selected for having high +- scores but on the amount PT they have received. Naturally Hornacek plays lineups that do well in his estimation so there indirectly they are selected for good +-.

But Thomas' and Green's +- ratings are so high they can't be riding on other guys coattails. Lets compare them in this manner with Marcus Morris who gets 48% of his minutes in there lineups and his +- in them is 2.8 vs 1.2 overall. The average +- for all 20 lineups is 8.8/48 so Marcus is dragging them down, while Green and Thomas are pulling them up.

I will spare you the gory details but Markieff remarkable -7/48 OFFfloor +- also follows from the fact that 83% of his PT comes from the top 20 lineups.

What are the lineups like that cover the remaining 970 or 45% of the team's minutes wherein we are outscored by 105 points? For starters none of those lineups play more than 18 minutes, so there at least 53 such lineups (970/18 = 53.9.) The average PT won't be as high as 18 so figure 12-13 minutes per, which gives 75-80 other lineups for a total of around a 100. That is quite a ridiculous number of lineups to juggle, especially with the total PT of most of them being so small you have almost no information about how worthwhile they are.
A quick look at 10 other teams show quite similar patterns - so NBA coaches all seem to be micro-managers. I'm going to have to look deeper but the way we fans talk about a second unit, as though was a reasonably well defined entity, seems not to exist on any team.
The funny thing is that I wrote the program I used to extract the info I used above when 82games.com first showed up, with the mistaken idea that the top 20 lineups would cover 90% of the teams minutes if not 100% - why would a coach need more than 20 lineups for crying out loud. When it turned the number was around 50% I didn't see much value in analyzing it. What sparked my interest today was that I was sure there had to be some explanation for Markieff's seeming pre-eminece in the +- stats when he clearly isn't a dominant force on the floor. And I wondered about Thomas and Green looking so bad in those stats, too.

My interpretation of why Markieff looks so good is, that first of all he's our best big. Secondly Jeff doesn't play him the heavy minutes you'd expect at only 30.5/game. He keeps him relatively fresh and only puts him on the floor with our other good players - and that may well be the best way to use him.

So here's a little bonus for those of you who read this far. How well do the twins work together? Not well at all, based on the top 20 lineups - when they're both on the floor their +- is 2.4/48; Markieff but no twin 13.5; Marcus untwinned 3.4. I looked at all pairs of players and there is no player that plays better with Marcus on the floor than he does without him. (Remember this is just based on the top 20 lineups so that could change if all the lineups were available.)

 

[Phrazbit]

With you missing nearly 45% of the overall minutes I'm not sure how much value you can pull from those stats. Especially if it shows Green as having such a great +,- when his overall figure in that stat is the worst of any rotation player on our roster.

You can find some of the deeper 5 man units if you go through the players individually. Each has is own individual top 20 rotations he has played in this year. Skimming through those might clear up some of those irregularities.

 

[Errntknght]

Even though I don't know how the 'missing' 45% breaks down I know how the team fares as whole - they are -105, so a player like Green who plays 63% of his minutes with those missing lineups has almost no chance of faring well statistically regardless of how well he does individually. Mentally I did the computation - if he played 600 of those 970 missing minutes his expected result would be 60% of -105 or -63 which would knock his +- down about where it wound up. I might go back and add in that computation because I know I'm not going to feel like gathering all the data very often and this computation is useful...

But I will try your suggestion of gathering up all those lineups from individual players and dumping them in one file. If they all extract to text the same way, its almost trivial to eliminate the duplicates and I should have all the lineups... its almost impossible for a lineup to not be in someone's top 20. Frankly, I don't remember what ideas I had for extracting information from the full set of lineups... I must have been going to look at pairs of players to see if there were pairs that did exceptionally well together because the program outputs all pairs now. And that is a case where you'd definitely want all the lineups. Lineups with under 20 minutes of data are pretty useless as lineups but a bunch of them are fine for information about pairs. You might wonder why I'd be interested in pairs of players... idle curiosity is about all I can think of. I'm rambling again...

 

[mojorizen7]

The not so advanced stats say...

-Allowing 105.1 ppg(3rd highest) and scoring 107.3(2nd highest)
-Allowing 12 offensive rebounds/gm(3rd most)
-Allowing 46.4 ppg in the paint(2nd most)

Yuck. The Song Remains the same...but hey they're winning in January right? What's to complain about?

 

[AzStevenCal]

The not so advanced stats say...

-Allowing 105.1 ppg(3rd highest) and scoring 107.3(2nd highest)
-Allowing 12 offensive rebounds/gm(3rd most)
-Allowing 46.4 ppg in the paint(2nd most)

Yuck. The Song Remains the same...but hey they're winning in January right? What's to complain about?

The main reason we are winning in January is that we've improved in these areas since moving Len into the starting lineup (mid-Dec). We still have a long way to go. Replace Green and Marcus with average defenders and out defensive stats start to look a lot better. Hopefully that's the plan and sometime soon.

Steve

 

[Covert Rain]

The not so advanced stats say...

-Allowing 105.1 ppg(3rd highest) and scoring 107.3(2nd highest)
-Allowing 12 offensive rebounds/gm(3rd most)
-Allowing 46.4 ppg in the paint(2nd most)

Yuck. The Song Remains the same...but hey they're winning in January right? What's to complain about?

Yeah but we have had other offensive minded teams win during the regular season. I was excited about this team building defense but as some others have pointed out...we have some defensive dead weights we need to ditch if this team is to make it to the next level IMO.

 

[AzStevenCal]

Yeah but we have had other offensive minded teams win during the regular season. I was excited about this team building defense but as some others have pointed out...we have some defensive dead weights we need to ditch if this team is to make it to the next level IMO.

I think that was his point. Regular season offense only = playoff failure. And, as you well know, he's right. Even though the league has changed quite a bit in the last several years, it hasn't changed that much. You still need to be able to defend and score in the half court. We have a lot riding on Len. He's already improved our defense with just glimpses of how much he'll be able to change it when he finally hits his stride (3 or 4 years?). He doesn't have much of a low post game right now but he's physical, has nice footwork, great touch and he's quick so there's no reason he won't be able to become a reliable half-court scorer. We just have to wait for it, unfortunately.

Steve

 

[Covert Rain]

I think that was his point. Regular season offense only = playoff failure. And, as you well know, he's right. Even though the league has changed quite a bit in the last several years, it hasn't changed that much. You still need to be able to defend and score in the half court. We have a lot riding on Len. He's already improved our defense with just glimpses of how much he'll be able to change it when he finally hits his stride (3 or 4 years?). He doesn't have much of a low post game right now but he's physical, has nice footwork, great touch and he's quick so there's no reason he won't be able to become a reliable half-court scorer. We just have to wait for it, unfortunately.

Steve

+1

 

[FArting]

We are 3rd worst in the league in giving up points.

Minnesota and Lakers are the bottom 2.

 

[Suns_fan69]

We also play at the 2nd fastest pace in the league. We're actually right smack in the middle at 15th at points per possession: http://stats.nba.com/league/team/#!/advanced/?sort=DEF_RATING&dir=1

 

[sunsfan88]

We are also the worst team in the NBA in pts allowed in the paint.