An Examination of The Mavericks' Best and Worst Five-Man Lineups.

All-Star Weekend has drawn to a close and, with that, the second half of the season has begun in earnest. So, let's get back up to speed.

Before the extended break came and went, you may recall that the Mavericks were in the midst of attempting to turn around their middling season through the powers gifted to them by facial hair. As it turns out, though, there may be something to the team's follicle freedom.

Dallas is 4-2 in the month of February. Looking ahead, the team only faces two opponents with winning records for the remainder of the month. Those teams are the Milwaukee Bucks and Memphis Grizzlies. The Mavs also play the Lakers before month's end but, unless you have been living in a yurt in the Kazak steppes, you should know by now that the Lakers have issues of their own.

With Dallas sitting six games below a .500 record, it is not out of the question that the players call their barbers and inform them to keep the straight razor on standby.

Still, nothing is a given with this team.

If the Mavs are to climb up to a respectable record, head coach Rick Carlisle is going to have to play is best units.

A couple of weeks ago, Ian Levy put together a chart at The Two Man Game, comparing how well a particular player pairs with another on the Mavericks. The results were quite interesting. The chart, which is based on Net Rating (point differential per 100 possessions), shows that Darren Collison and O.J. Mayo make for a particularly poor pairing,- and yet, here they are, the starting backcourt for Dallas.

Dirk Nowitzki also does not fare much better when on the floor with either Collison or Mayo. Perhaps most surprising about Levy's chart: It shows that Shawn Marion, who is undoubtedly one of the few players who has remained consistent throughout the season, is bad when paired with just about any player on Dallas' roster.

These four starters — now that Nowitzki is healthy — make up the core of the starting rotation. Nonetheless, one look at that chart shows it: They do not play well together. And, generally speaking, it's not good when the starting rotation plays poorly together. Not good at all.

This look at the team is limited, of course. It only takes into account player pairings — not how the whole unit, all five players on the court, plays together.

But thanks to NBA.com embracing the advanced stats revolution that has swept through the league over the past decade or so, anyone can now access various statistics pertaining to almost all aspects of basketball.

In this case, let's look at the units that the Mavericks have employed this season.

Dallas has placed 251 different units on the court this season. For the purposes of this post, however, I will only look at units that contain players who are currently on the roster. (Sorry, Derek Fisher.)

The lineups that have received the most minutes this year are:
• Elton Brand, Vince Carter, Collison, Marion, Mayo (Unit 1)
• Carter, Collison, Chris Kaman, Marion, Mayo (Unit 2)
• Collison, Kaman, Marion, Mayo, Nowitzki (Unit 3)

Each of the first two units have played a total of 135 minutes together. The third unit has played a total of 125. (If you're wondering why Nowitzki is not present in the two units with the most minutes played, you have to remember that he missed much of the beginning of the season recovering from surgery on his knee.)

Using the same metric as Levy, the first unit (Unit 1) has a Net Rating of 20, while the second unit (Unit 2) has a rating of 16. Scoring either 20 or 16 more points than your opponent per 100 possessions is quite good. Interestingly, this is despite Unit 1 turning the ball over 15.9 percent of the time (leading to 64 points scored by opponents) and Unit 2 doing so at a clip of 14.7 percent (with 47 points scored against them). Where these two units benefit is their high Effective Field Goal Percentage, a percentage that takes into account a three-point field goal being worth more than a two-point field goal.

While it is clear to see why these two units have been played as much as they have, the third unit (Unit 3) fares a slightly different fate. Unit 3, the standard starting five for Dallas since Nowitzki's return, has a Net Rating of -34. That is not good. And in some ways gives validity to the argument that Kaman and Nowitzki should never be on the court at the same time, given their similar shot selections and sub-par defensive capabilities. However, if one were to substitute out Kaman from this unit and replace him with Brand — a unit that has played 61 minutes together — the Net Rating would be exactly the same (-32).

Recent weeks have found the Mavs playing a new starting center, though, in Bernard James. He began receiving the starting nod a few days before Kaman suffered a concussion in practice. Though James complements Nowitzki's style of play by playing around the basket, this unit, also containing Collison, Marion and Mayo, which has played 68 minutes together, also has a Net Rating of -10.

In fact, the only lineup that features Collison, Marion, Mayo, and Nowitzki that has a positive Net Rating is one featuring Brandan Wright at center. That unit has a Net Rating of 4 and has only played together for six minutes.

Though Levy's initial findings were somewhat limited, they also noticed a larger trend. The four players that the Mavericks have featured most in their starting lineup do not play well together — not statistically, at least.

Perhaps it's time for Carlisle to completely rethink his starting rotation. It will take some thought on his part, of course. His best lineups to date have been the ones without Nowitzki, and you know Carlisle's never going to bench him. Nor should he, necessarily.

But, at this point, he's got to do something.

Cover photo via the NBA's official Mavericks page.

3351_2

3351_3

3351_4

3351_5

3351_6

3351_7

3351_8

3351_9

3351_10

3351_11

3351_12

3351_13

3351_14

3351_15

3351_16

3351_17

3351_18

3351_19

3351_20

3351_21

3351_22

3351_23

3351_24

3351_25

3351_26

3351_27

3351_28

3351_29

3351_30

3351_31

3351_32

3351_33

3351_34

3351_35

3351_36

3351_37

3351_38

3351_39

3351_40

3351_41

3351_42

3351_43

3351_44

3351_45

3351_46

3351_47

3351_48

3351_49

3351_50

No more articles
X