Shifty

Posted by Ian Levy on March 29, 2012 under Commentary | Read the First Comment

Screen Shot 2012-03-29 at 9.56.53 AM

When you see measures of a team’s pace, it’s almost always listed as a single number – the average number of possessions a team uses per game. The addition of pace into the statistical lexicon was a monumental step forward, but discussions on the subject still lack oodles of nuance. The Mavericks are averaging 91.7 possessions per game this season, but those 91.7 possessions are not consumed in an equal distribution across 48 minutes. The game speeds up and slows down, alternately racing and lulling through to its conclusion.

Looking at pace only in the context of average possessions used per game leaves a lot of information by the wayside. For example, a team might play two games, using 90 possessions in each, for a pace factor of 90. Another team might play two games, using 70 possessions in the first and 110 in the second, also arriving at a pace factor of 90. Those two pace factors, although numerically identical, are far different in terms of functional significance. These fluctuations in pace don’t just happen across single games, but also across sections of games; teams will play at different speeds based on situations or the personnel that’s on the floor, all of which is muddled in a single pace factor.

Last season I tried to dissect pace in two slightly more defined ways. During the playoff series against the Thunder I looked at pace in smaller chunks of time, and found that the Mavericks seemed to struggle as the game got faster. I also looked at how the Mavericks pace changed when different lineups were on the floor.

What I was curious to look at next was how much the Mavericks’ pace fluctuated in settling on their average, and how that compared to the rest of the league. The tool I used to illustrate this change is variance, a numeric expression of how much a data-set varies from the average. For this analysis I looked at the ten most played lineups for each NBA team and calculated the variance for those ten lineups from the team’s average. The higher the variance, the more change there was in pace from unit to unit, the smaller the variance, the more consistent their pace was.

PaceVariance

Note: My data comes from Basketball Value, so unfortunately I don’t have numbers for the Memphis Grizzlies.

Here’s the same information in numeric form:

Team ▴ORtgDRtgNetPacePace Variance
Memphis Grizzlies101.9100.51.491.6-
Minnesota Timberwolves103.1101.91.293.536.4
Denver Nuggets106.9104.72.294.730.2
Boston Celtics100.199.11.090.329.7
New Jersey Nets103.9111.5-7.689.428.7
Dallas Mavericks102.198.33.891.724.5
San Antonio Spurs107.3103.34.092.221.7
Charlotte Bobcats94.4110.0-15.691.319.6
Golden State Warriors106.7108.3-1.692.219.0
Toronto Raptors98.4104.0-5.690.017.5
Chicago Bulls108.297.710.589.616.0
Phoenix Suns102.6105.2-2.692.115.4
Los Angeles Lakers103.0101.02.090.314.2
Portland Trail Blazers105.499.65.891.413.7
Miami Heat109.9100.09.991.813.6
Cleveland Cavaliers101.4105.3-3.991.813.3
Detroit Pistons100.1107.5-7.489.012.2
Milwaukee Bucks102.6105.0-2.492.911.0
Los Angeles Clippers107.9105.02.988.910.4
Oklahoma City Thunder108.9102.16.893.310.0
Sacramento Kings100.3108.8-8.594.309.4
Indiana Pacers103.8100.63.290.808.7
New York Knicks100.899.61.294.108.1
Houston Rockets105.0103.31.791.707.9
Utah Jazz104.6106.3-1.791.507.5
Atlanta Hawks101.9100.11.889.806.7
New Orleans Hornets99.1104.8-5.788.506.6
Philadelphia 76ers104.297.07.289.804.8
Washington Wizards97.9107.4-9.592.904.5
Orlando Magic103.4100.92.589.403.3

Very good teams are scattered throughout this list, and there certainly don’t seem to be global patterns of connection between pace — or variability of pace — and winning games. The Mavericks’ pace of 91.7 this season falls right around the league average. However, among their 10 most used lineups, the Mavericks have had the 5th most variable pace. These numbers have changed over the course of the season, but every time I’ve looked through the data, the Mavericks have been towards the top. I’m shifting towards hypothesis here, but it seems that lineup fluctuations could be indicative of three scenarios – a team whose identity is not tied to pace and therefore has that characteristic defined by their opponents, a team that has an identity tied to pace but does not have the talent or execution to dictate that pace on a consistent basis, or a team that purposefully focuses their personnel on game situations and is comfortable with their ability to compete at any speed. Assuming that those three categories are at least somewhat indicative of reality, I think it’s safe to say the Mavericks fall comfortably into the third.

During the playoffs last season, Rick Carlisle revealed himself as a master puppeteer, pulling strings with fingers and toes, manipulating matchups and extracting the very best from the roster he was given. Without the benefit of continuity or consistent health, the level of difficulty on this task went up significantly this season. Looking at those 10 most used lineups clears some of the haze and thrusts those difficulties forward.

PlayersMinOff Rtg.Def Rtg.Net Rtg.PaceDiff. From Season Average
Kidd - Carter - Marion - Nowitzki - Haywood153.85112.0386.4425.5991.4-0.3
Kidd - West - Marion - Nowitzki - Haywood130.3597.1793.124.0591.0-0.7
Kidd - Terry - Marion - Nowitzki - Haywood98.08113.44102.710.7490.8-0.9
West - Carter - Marion - Nowitzki - Haywood76.92122.4583.5538.9093.31.6
Kidd - Carter - Marion - Nowitzki - Mahinmi 53.58102.83100.931.9095.94.2
Terry - Carter - Odom - Nowitzki - Mahinmi 49.83129.76106.0223.7480.4-11.3
Beaubois - Terry - Odom - Nowitzki - Mahinmi40.97112.16122.86-10.7084.4-7.3
Kidd - Terry - Marion - Odom - Mahinmi40.5797.50113.42-15.9195.84.1
Kidd - Terry - Marion - Nowitzki - Mahinmi38.5791.5585.516.0487.1-4.6
Beaubois - Terry - Odom - Nowitzki - Wright38.08122.37100.0022.3793.92.2

First off, in those 10 lineups, you’ll find 11 different players, a testament to how much injuries have affected the Mavericks’ rotations this season. Second, it’s obvious what a challenge it has been to find a place for Lamar Odom. Dirk Nowitzki and Shawn Marion are both central pieces to what the Mavericks hope to accomplish at both ends of the floor and thus you would expect to often find them occupying the two forward slots. But the fact that none of their five most used lineups feature Odom is a testament to how poorly he’s played and how poorly he’s been utilized. It would probably be much easier for Carlisle to write him off, but Odom is so talented and still has the potential to help the team in so many ways. Subtly pushing him down the bench and out of mind simply isn’t compatible with the goal of competing for a second championship — hence why he’s managed to linger in the rotation even through the worst of his struggles.

The good news is that in these units, there is at least one scenario in which Odom has been very successful. The scenario I’m referring to is his place in the most numerically striking unit in our data: Terry – Carter – Odom – Nowitzki – Mahinmi. This group is playing at a pace 11.3 possessions per 48 minutes slower than the Mavericks’ team average. The amount by which this unit deviates from the team pace is extremely rare – just 13 lineups in our data set for the entire league varied from their team averages by more than 8 possessions:

TeamPlayersMinOff Rtg.Def Rtg.Net Rtg.PaceDiff. From Season Average
DALTerry - Carter - Odom - Nowitzki - Mahinmi49.83129.76106.0223.7480.4-11.3
BOSRondo - Allen - Pietrus - Pierce - Garnett36.30116.67127.40-10.73103.813.5
CHAAugustin - Henderson - Maggette  - White  - Diaw89.6098.40113.44-15.0599.98.6
DENLawson - Miller - Afflalo - Harrington - Hilario 39.9297.22124.24-27.0283.0-12.1
MINRidnour - Ellington - Johnson- Love - Milicic48.8889.41120.00-30.5983.5-10.0
MINBarea - Webster - Beasley - Williams - Milicic33.28136.84100.0036.8483.7-9.8
MINRubio - Barea - Beasley - Williams - Love32.3392.6590.002.65102.48.9
NJNWilliams - Stevenson - Brooks - Humphries - Okur35.05108.62110.17-1.5580.1-9.3
NJNGaines - Brooks - James - Williams - Okur17.92106.67103.453.2279.0-10.4
PHXNash - Brown - Dudley - Frye - Gortat40.7089.29106.98-17.69100.28.1
PORCrawford - Matthews - Batum - Wallace - Aldridge57.1397.5299.15-1.62100.08.6
SASParker - Neal - Green - Bonner - Splitter40.52130.88114.4916.3981.1-11.1
TORCalderon - Barbosa - DeRozan - Davis - Johnson35.53112.90101.7211.1881.1-8.9

That group has also been one of the Mavericks’ most successful units this season, outscoring opponents by 22.74 points per 100 possessions. With an Offensive Rating of 129.76, the strength is efficient scoring. A look through NBA.com’s new statistical database reveals that this lineup is exceptional in their ability to create quality three-point looks. Of the league’s 250 most used five man units, this group has the 13th highest 3PT%, at 51.9%, with 92.9% of those makes being assisted on.

I don’t have answers for why this unit has been so uniquely successful with Odom, when others have seemed to struggle so mightily. I think pace is playing a big role here, but it’s as much a symptom as a cause. This unit is playing very slowly, and scoring at a very high rate — a combination which indicates execution in the extreme. It also seems that with Carter and Nowitzki, two post scoring threats, and Terry, a potent spot-up shooter, this unit probably puts Odom in a role most similar to the one he played in Los Angeles.

Throughout his career, Odom has been a terrific offensive player, but he has always been most successful as a complementary one. In Dallas, he has been asked to use his complementary skill set in way that is largely new to him. Odom is averaging a career high 3.9  three-point attempts per game with an almost career low 13.3 Ast%. Being used so frequently as a spot-up shooter (often due to his own lack of movement) does not play to Odom’s strengths – a big man who can handle the ball, attack an out-of-balance defense, and find open teammates. On some level, it seems like he is stepping into the box that Peja Stojakovic filled last season, one that is a terrible fit for him. His generally miserable play and lack of engagement haven’t helped any, but perhaps part of the solution is asking him to do some different things on the floor.

One other interesting note: in looking at other player combinations among these units I also noticed this was the only time Odom and Carter appeared together. I again turned to NBA.com’s new stats tool and found that the Odom and Carter pairing has resulted in the Mavericks being +4.8 per 48 minutes. Considering that the Mavericks have been -1.9 per 48 minutes with Odom on the floor as a whole, that number is pretty significant. Just 35.2 percent of Odom’s minutes have been played alongside Carter, but exploring that pairing may be another piece of the solution.

The Terry – Carter – Odom – Nowitzki – Mahinmi unit stands out when compared to the team and the rest of the league. It is a unit that is as deliberate as any in the Mavericks repertoire or that can be found around their league, yet Dallas has shown the ability to create and make quality open shots within that configuration — the engine that drove the Mavericks’ championship run last season. Pinpointing the whys and hows, could help replicate and spread that success.

The rough pace data for every team’s lineups can be found here.

Statistical support for this story from NBA.com.

Ian Levy is the author of Hickory High, a contributor to Indy Cornrows and HoopSpeakU, and a part of The Two Man Game family. You can follow Ian on Twitter at @HickoryHigh.

  • Jensen Matlock

    It's surprising that that lineup would be so efficient without a primary ball-handler.  Odom is certain capable of filling that role which Rick had him doing early on in the season but has since gone away from it due to Odom's poor play and lack of offensive flow.  The slower pace makes sense without Kidd or Roddy or West pushing it up the court and without someone like Marion who loves to get out on the break.  The efficiency is certainly the surprising part (especially since I doubt that lineup is ever together during practice).  Clearly there is something in that lineup that makes it easier for Lamar to be effective or at least not a detriment.

    I've been disappointed the last couple games that Rick put Roddy back on the bench cause I think he can give us a good jump at the start of the game but also because I like the idea of having a 6a and 6b like we had last season with Terry and Marion before Butler went down. Because they were set up in the rotation to play together, they practiced together all the time and I thought that gave them great chemistry on the court – early in the season the passing between them was awesome.  The same thing could work with Odom.  Rick has done a great job of being patient with Odom and trying him in different spots and different roles but maybe a more consistent lineup pairing (maybe Carter with Roddy starting) would help him develop some chemistry and get into the offensive flow.  Just a thought.