Good and Bad Offensive Lineups

When I was in sixth grade, my basketball coach had these words to say after a particularly disheartening first half of play:

"Just put the ball in the basket.  Don't worry about anything else.  Just take this ball (holding it high above his head) and put it inside the rim."

Well.  The NBA Season is over 1/3 completed and my coach's advice is still some of the best I've ever heard.  It is time to consider who is good at putting the ball in the bucket and who just...isn't.  Let's look at the top 5 offenses in the NBA, per NBA.com:

1.  Portland Trail Blazers (110.1 Points Per 100 Possessions)

2. Miami Heat (109.1)

3. Houston Rockets (107.4)

4. San Antonio Spurs (107)

5. Oklahoma City Thunder (106.6)

And the Bottom 5:

30. Milwaukee Bucks (96)

29. Chicago Bulls (96)

28. Charlotte Bobcats (96.2)

27. Cleveland Cavaliers (97.4)

26. Utah Jazz (97.7)

What do the top teams have in common?  The bottom teams?  Let's just say their teams' front offices will be focusing on much different things come May-June.  Although defense is very important, scoring the ball remains vital to team success in the NBA.  The Bobcats boast the 6th ranked defense, but remain under .500 because of their offensive problems (even in the rotting carcass that is the Eastern Conference).

Some teams, although they are overall very good (or bad) offensively, experience large gains or drop-offs depending on which 5 man lineup they use.  For example, let's look at the Bobcats again.  Here are a couple of 5-man offensive lineups that Coach Clifford gives a lot of minutes:

Lineup 1: Kemba Walker- Gerald Henderson- Anthony Tolliver- Josh McRoberts- Al Jefferson

Vs.

Lineup 2: Kemba Walker- Gerald Henderson- Michael Kidd-Gilchrist- Josh McRoberts- Bismack Biyombo

Which lineup would you guess knows how to put the ball in the basket?  Lineup 1 scores 115 points per 100 possessions.  Lineup 2 scores just 89.  Judging from the individual players' reputations, this is not a surprise. You probably guessed the first lineup was better offensively.  You would also be correct in assuming the second is better on  the defensive end (DRTG of 104 to 94).    

Can we explain the huge drop-off in offense from Lineup 1 to Lineup 2 for the Bobcats?  What about different lineups for all NBA teams?  

I analyzed the top 90 NBA lineups in terms of Minutes played together so far in the season.  I gathered the following data from NBA.com's player tracking data for each player in these lineups (each on a per-minute basis):

• Close Touches
• Time of Possession
• Frontcourt Touches
• Elbow Touches
• Assists
• Drives
• Catch and Shoot FGA
• 3 Point Catch and Shoot FGA
• Off Dribble FGA
• 3 Point Off Dribble FGA

I then checked to see which of these factors, sorted by position, played the biggest role in determining a high Offensive Rating.  And here are the findings:

I included the heads of players from good offenses that exemplify these characteristics.  

Some Notes/Observations/Interesting Tidbits/Debatable Topics:

• There is evidence here to suggest that a "Stretch Four", as many call it, is important, if not necessary, to a successful NBA offense.
• It is more important for a 2 Guard to drive than a Point Guard
• Small Forwards are most effective as ball movers and threats from beyond the arc.  
• Even with the demise of the traditional center, offenses still depend on their biggest player to shoot close to the basket.
• Shooting off the dribble is one of the most skilled things an offensive player can do.  Having skilled power forwards and shooting guards helps an offense.

And here are the findings for bad offenses: 

(Note: I didn't include pictures this time, because I don't want to hate on any players.  Also my Utah Jazz friends wouldn't have liked me associating Gordon Hayward's head with anything negative.)

Some Notes/Observations/Interesting Tidbits/Debatable Topics:

• The negative impact of a Shooting Guard's Catch and Shoot really jumps out here.  It speaks to how much the "Shooting Guard" has really turned into more of an "Attacking Guard".  
• As discussed before, a Small Forward is better served either catching and shooting or moving the ball rather than shooting off the dribble.
• Power Forwards help their team the most when their are stretching the defense and finishing rather than facilitating.
• Centers that try to do too much are hurting their team (or maybe the current crop of centers just isn't that talented?)
• Point Guards should not be in the paint.  

Got a comment?  Question?  Suggestion on how to improve the study?  Please comment.

(A couple of disclaimers: I also ran regressions on different combinations of these factors, and nothing was statistically significant.  It is easy to see why; there is just so much variation between players and unexplained reasons why certain lineups perform better.  So although these factors have higher correlation to offensive rating, this is in no way conclusive.  Also, some players were more represented than others in the analysis, because of playing time/coaching decisions/injuries to themselves and teammates.  For example, John Wall's factors may be over represented, as he is part of 5 of the top 80 lineups in total minutes played.)