IT-and-D problem? Boston Celtics' three-guard lineups have been getting shredded

http://www.masslive.com/celtics/index.ssf/2016/12/it-and-d_problem_boston_celtic.html



IT-and-D problem? Boston Celtics' three-guard lineups have been getting shredded





Boston Celtics' Isaiah Thomas (4) goes up for the go-ahead shot past Philadelphia 76ers' Dario Saric (9) and Jahlil Okafor ( during the final minute of an NBA basketball game, Saturday, Dec. 3, 2016, in Philadelphia. Boston won 107-106. (AP Photo/Matt Slocum)



Jay King | mjking@masslive.com By Jay King | mjking@masslive.com




on December 04, 2016 at 6:00 AM






Boston Celtics head coach Brad Stevens called on one of his favorite lineups late in the fourth quarter Saturday night, and a shootout ensued. The Celtics prevailed thanks largely to Isaiah Thomas' affinity for the rim, but still might have real concerns with their "IT-and-D" unit: Thomas, Avery Bradley, Marcus Smart, Jae Crowder and Al Horford.

In that group, Stevens knows what he has: a potential offensive rocketship that only needs to hold its own at the other end of the court. The problem so far has been that the Celtics defense has gotten shredded with that lineup, as well as with most lineups that include Thomas, Bradley and Smart. Theoretically, that trio could be great. Smart can guard anyone on the perimeter, Thomas destroys opponents almost nightly, and Bradley, a first team all-defense member, adds to his game every year. But reality has produced a distressing lack of stops and defensive rebounds when those players -- who are all good! -- share the court.

Take the final five minutes against the Philadelphia 76ers, when the Celtics needed Thomas to score a dozen straight points just to keep pace with maybe the NBA's worst team. Over the 4:30 when the Celtics had the IT-and-D lineup on the court, the 76ers went 4 for 7 from the field, including 2 for 4 from behind the arc, and rebounded two of their three misses. Down two in the final minute, they found power forward Dario Saric guarded by the 6-foot-2 Bradley in the post:



Even when the shot clock hits one -- and Saric desperately needs to shoot regardless of how he's defended -- the Celtics fail to give any help. They are probably worried about a possible rebound, but never reach that step because Saric takes advantage of the mismatch with a drop-step to the baseline.

Since Horford and Crowder returned from injuries, Stevens has frequently used the IT-and-D lineup to close out halves. It has gotten thwacked too often. The sample size for that group is tiny (36 minutes played), but consider this: among all lineups across the NBA with at least 35 minutes, only one has a worse defensive rating. One of the Celtics' go-to crunch-time lineups is getting bundled on the glass and ransacked at the 3-point arc. It has been great offensively, which speaks to the untapped potential, but needs to transform itself into at least a passable defensive unit.

"Obviously, it's a little bit of a catch-22 because of their size, right?" Stevens said after the San Antonio game, when asked how that lineup can get more stops. "So you've got a lineup that can really space people out, drive the ball, (you) saw Al gets a couple of rolls to the basket and the basket is open for him. At the same time, like (against the Spurs), we didn't rebound on two of the possessions in the last two minutes. I think that those guys have all proven themselves to be amongst our better players on this team. And, at the same time, we're going to have to figure out how to manage their size a little bit better, especially against guys like (LaMarcus) Aldridge."

Why does Stevens keep going to the lineup? Well, it has five of his best players, including two All-Stars and a first team all-defense member. When Smart handles the ball with four threats around him, the Celtics reach peak floor spacing:


via GIPHY

And when Thomas has the ball with skill everywhere around him, good luck trying to keep up:



"He's really good. He's really, really good," Stevens told reporters in Philadelphia. "My advice to anybody is to talk whoever you work for into buying League Pass and watch him as much as you can. Because he's a fun guy to watch."

It's not crazy to envision a day when Boston's super-small lineups force mountains of turnovers, create loads of transition offense, and ride down a slide of layups and 3-pointers. But right now, that's all just a dream. While it's too early for Stevens to stop trying to make things work, the Celtics defense got diced when Thomas, Bradley, Smart and Crowder shared the court last season, too. While the offense was still powerful enough to make that quartet a net positive, the problems -- especially rebounding -- have worsened this season, and Boston hasn't forced nearly as many turnovers. With Evan Turner gone, Stevens doesn't trust as many other lineup options. He has been bullish about the super-small groups, and they haven't worked yet.

So far, the Thomas-Bradley-Smart combination has played 252 minutes together, which equates to more than one-fourth of Boston's season overall. During that time, the Celtics have been outscored by 6.3 points per 100 possessions. Essentially that means when three of their best players share the court, they have played basketball like the 4-15 Dallas Mavericks, with a downright appalling 112.0 defensive rating.

Knowing that number, it's easier to understand why the Celtics once-proud defense currently ranks 18th. The starting lineup has been dynamite and the offense is overhauled from last season, but Stevens is still searching for the right complementary groups. Sitting alone in third place in the Eastern Conference after an injury-riddled start is nice, but the Celtics might have another level to reach.

Last year, lineups with Thomas, Bradley and Smart actually outscored opponents by 5.4 points per 100 possessions, but the defensive rebounding rate for those groups has dipped from steady to catastrophic. All year long, rebounding survival should be a theme. Bradley has crashed the glass like a mad man, but it's ominous to have a 6-foot-2 leading rebounder. Turner and Jared Sullinger helped keep the Celtics afloat on the glass last season, and Danny Ainge neglected to find another physical presence down low. For everything Horford brings, rebounding is just about the one thing he doesn't do at an above-average level.

"We're giving up a lot of second-chance points when we get that (three-guard) lineup," Crowder said last week. "Scoring is not a problem with that lineup. We just have to find a way to get stops. And that's by rebounding that ball. All five guys need to be on the same page, and just try to end the possession.

"(Switching) puts a guard on a big. And that's just him having to take the big out of the play. We have to be more aware of when a smaller guard is on a big and chop down with two, because most teams don't send four or five guys to the boards. So we can have all our guards help out when Isaiah or Avery is on a big or something like that. We just have to be more cautious of it."

Almost one-quarter of the way through the regular season, Stevens' team is healthy and on pace for 49 wins. That's not too far off reasonable expectations, but the Celtics will only reach their potential if they find a way to thrive when all their best players share the court.




bob



.

And what about the other teams. It is not like other teams can stop our guards.

dboss

From the numbers at "nba.com/stats league player tracking defense" it appears that much of the blame can be placed at the feet of Bradley. Using the latest in stadium video recording they now record every shot taken - from where - who took the shot - against whom and from the distance taken and of cause made or missed. This enables them to compile each defensive players defensive shooting % against the player they were guarding. In addition they have yearly stored data for each player being guard. They then compare the defensive players fg % with the offensive players yearly shooting %.

Note" the lower the difference the better the rating.

Bradley,,,, players guarded by him are shooting .486 - guarding the rest of the league at .433 - diff plus 5.3
Smart,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,.438 - ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, .446 - ,,, minus.0.8
Thomas,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,.410 - ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, .439 - ,,,, minus 2.9

http://stats.nba.com/league/player/defense/#!/?Season=2016-17&SeasonType=Regular%20Season&PlayerPosition=G&sort=PCT_PLUSMINUS&dir=-1

Plenty of sites with offensive stats but they are quite rare on the defensive side. I've been searching for a long time and finally hit this one the other day. When you pull up the Celtic players you can get all the details for the various distances by clicking on to their name - Note: I sorted by the guard position.

swish

swish wrote:From the numbers at "nba.com/stats league player tracking defense" it appears that much of the blame can be placed at the feet of Bradley. Using the latest in stadium video recording they now record  every shot taken - from where - who took the shot - against whom and from the distance taken and of cause made or missed. This enables them to compile each defensive players defensive shooting % against the player they were guarding. In addition they have yearly stored data for each player being guard. They then compare the defensive players fg % with the offensive players yearly shooting %.

   Note" the lower the difference the better the rating.

  Bradley,,,, players guarded by him are shooting .486 - guarding the rest of the league at .433 - diff plus 5.3
  Smart,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,.438 - ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, .446  - ,,, minus.0.8
  Thomas,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,.410 - ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, .439 - ,,,, minus 2.9

http://stats.nba.com/league/player/defense/#!/?Season=2016-17&SeasonType=Regular%20Season&PlayerPosition=G&sort=PCT_PLUSMINUS&dir=-1

 Plenty of sites with offensive stats but they are quite rare on the defensive side. I've been searching for a long time and finally hit this one the other day. When you pull up the Celtic players you can get all the details for the various distances by clicking on to their name - Note: I sorted by the guard position.

  swish  

swish,

Your analysis supports the article's premise, that our presumed best defensive 5, except for IT, are not doing as well as some peoples' eye-tests say they are.  Our overall defense is a not-so-good 18th in Defensive Rating (points allowed/100 possessions).  Interestingly, we are 14th in points allowed/game and 14th in giving up paints in the paint.  We are also 14th in points scored/game.  So, what is the cause of our 18th Defensive Rating, if all the other factors are seemingly locked in on 14th?

Possibilities:

1.  We lead the league in 'loose balls recoved'.  So, when it pops loose, for whatever reason, we're on it like a bird on a June bug.http://stats.nba.com/league/team/#!/hustle/?sort=LOOSE_BALLS_RECOVERED&dir=1

2.  We are 2nd in 'charges drawn'.  That's an empty possession for them.  Who are the top charge takers on the Celtics?  Kelly is 2nd in the league in charges drawn, with .5/game (Ilyasova with .6, Bogut with .5) but is tied with Ilyasova at .022 charges drawn/minute.  Smart is 2nd on the Celtics with .29, Crowder is 4th with .17 and IT 5th with .15.  Kelly is not in the "IT and D" lineup, but the others are.  Overall, however, his numbers are included in the team's defensive performance and rating.
http://stats.nba.com/league/player/#!/hustle/?sort=CHARGES_DRAWN&dir=1&Season=2016-17&SeasonType=Regular%20Season&TeamID=1610612738

Other considerations:

1.  Top Celtics for 'deflections' are Jae Crowder, Avery Bradley, Marcus Smart, Al Horford (Kelly is 5th).  There's the 4 'D' in the 'IT and D' lineup.  http://stats.nba.com/league/player/#!/hustle/?sort=DEFLECTIONS&dir=1&Season=2016-17&SeasonType=Regular%20Season&TeamID=1610612738

2.  Top Celtics for 'loose balls recoved' are Jae Crowder, Avery Bradley, Marcus Smart, Al Horford.  There's the 4 'D' in the 'IT and D' lineup.  http://stats.nba.com/league/player/#!/hustle/?sort=LOOSE_BALLS_RECOVERED&dir=1&Season=2016-17&SeasonType=Regular%20Season&TeamID=1610612738

3.  Top Celtics for 'contested 3pt shots' are Jae Crowder, Al Horford, Marcus Smart, Avery Bradley and IT.  There's all 5 of the 'IT and D' lineup.  http://stats.nba.com/league/player/#!/hustle/?sort=CONTESTED_SHOTS_3PT&dir=1&Season=2016-17&SeasonType=Regular%20Season&TeamID=1610612738

So, there are stats backing up the article's premise.

On the other hand, we are 17th in 'contested shots', specifically, we are 20th in 'contested 2pt shots'.  We are giving up too many open, easy shots.  Interestingly, Brooklyn is 5th in contested 2pt shots and 2nd in contested 3pt shots.  Overall, the Nyets are 2nd in the league in contested shots, and yet they stink.  So, maybe that isn't a key stat?  They also score more points/game than we do.  They are the 3rd worse team in the league while we are 9th best.  

My point?  I'm not sure I have one so much as an observation:  stats don't always explain things and when they do, they don't always explain them well.  "There are 3 kinds of lies:  lies, damn lies and statistics" - Mark Twain.   Statistics are good for supporting the eye-test, and vice versa.  the eye-test should be persuaded by statistics.  For example, I see Kelly taking a lot of charges, so I'm inclined to think well of that part of his game.  Would I have guessed that he'd lead the team and be near the top of the league leader board in that statistical category?  No, but now that I have seen that it comes together a bit and is put in perspective against his peers.

We can always count on you doing your homework, Swish.


bob


.

bob

And here's a peek at how the centers are doing in respect to DEFENSIVE field goal efficiency. Might come in handy when the trade talk starts to heat up. I'm quite surprised by some of the numbers - which is the reason that I prefer to back up the eye test with some numbers. My eyesight tends to be quite biased on occasion. Click onto each players name to get the full impact of how he fared at different distances.

Note: It will be a good idea to get familiar with all the terminology so that you can utilize all the sorting capabilities etc. such as the GEAR.

swish

swish wrote:bob

 And here's a peek at how the centers are doing in respect to DEFENSIVE field goal efficiency. Might come in handy when the trade talk starts to heat up. I'm quite surprised by some of the numbers - which is the reason that I prefer to back up the eye test with some numbers. My eyesight tends to be quite biased on occasion. Click onto each players name to get the full impact of how he fared at different distances.

  Note: It will be a good idea to get familiar with all the terminology so that you can utilize all the sorting capabilities etc. such as the GEAR.

 swish

Swish,

You forgot to include the link.


bob


.

bob

And what good is that post without the link?

http://stats.nba.com/league/player/defense/#!/?sort=PCT_PLUSMINUS&dir=-1&Season=2016-17&SeasonType=Regular%20Season&PlayerPosition=C

swish

swish wrote:bob

 And what good is that post without the link?

 http://stats.nba.com/league/player/defense/#!/?sort=PCT_PLUSMINUS&dir=-1&Season=2016-17&SeasonType=Regular%20Season&PlayerPosition=C

swish

Swish,

Interesting.  So, Kelly has a -1.0 differential (fg% of players he's defending - season fg%) but KAT is a +2.3.  I think that just highlights the dangerous weakness of relying on stats rather than many years of eyeballs.  CLEARLY, Karl-Anthony Towns is a better player than Kelly Olynyk, however, you wouldn't know it if you were to focus on this stat.  As I have been saying about Marcus Smart, there are two sides of the basketball court and your value to the team is the net of what you do to help the team on both sides.

Speaking of Marcus Smart, if you change the sort criteria from 'center' to 'guards', Smart has a -.7 differential.  That is worse than Monta Ellis (-1.2), worse than Rajon Rondo (-2.1), worse than Steph Curry (-2.6) and worse than teammate Isaiah Thomas (-2.eight).  Does anybody believe that IT is a better defender than Smart?  I sure don't and I don't know anybody on this board who has been more critical of Smart than me.  This analysis is defensive differential only and doesn't take into account IT's offensive contributions neither.  

I question these statistics.  


bob


.

bob

If I could have sorted by min per game Kelly's name would not have been listed ( short min - 21 per game ) Ditto for Zeller and all the other short minute players on the Celts.
Note: In below link notice how Thomas dominated Smart at 15 ft and out - while Smart dominated Thomas at 15 ft and in. Perhaps Smarts bulk makes his perimeter defense a little suspect - whereas Thomas'es quickness works in his favor - and his lack of size puts him at a disadvantage in the paint.

http://stats.nba.com/player/#!/203935/defense-dash/?p=marcus-smart
http://stats.nba.com/player/#!/202738/defense-dash/?p=isaiah-thomas

swish