Statistical look at Sixers starting lineups

Heading into last night's game against the Lakers, the Sixers had gone 5-2 over their first 7 games since inserting Jodie Meeks into the starting lineup, with Doug Collins seemingly settling in on both a starting lineup and a set rotation.  The lineup has been successful.  In fact, as John Schuhmann pointed out, the lineup is second in the league among lineups that have played at least 100 minutes together.

Some of that is reflective of the schedule the Sixers have played.  The 5 wins have come over poor teams (Clippers), struggling teams (Hornets, Bobcats), or both (Nets, Cavs).  The ease of the schedule has also been aided by the Sixers playing mostly at home, as 5 of the 7 games -- and 4 of the 5 wins -- have come at the Wells Fargo Center.  It would take a minor miracle for the lineup to maintain this kind of effectiveness over the next 9 games.

Still, it's hard to deny the lineups effectiveness.  Even against the legitimately good teams the lineup has done well.  Both the Sixers losses you could make a case they should have won, and the lineup held their own against the very strong starting lineups of the Boston Celtics and Atlanta Hawks.  The starting lineup was +12 during its time on the court against the Hawks and +6 against the Celtics.

For perhaps the first time this year Doug Collins appears set on his starting lineup and core rotation, having used 5 different starting lineups in the first 18 games of the season before the latest change.  The power forward/center combination of Elton Brand and Spencer Hawes has gone virtually unchanged, with the lone exception being when Thaddeus Young replaced Elton Brand after Brand received his one game suspension for his flagrant foul on JaVale McGee.  There was also the time Evan Turner took over for the injured Andre Iguodala.  

For the most part, the lineup changes have been an ongoing effort to find the right player to pair between Jrue Holiday and Andre Iguodala on the perimeter.  Jason Kapono, Andres Nocioni, Evan Turner and now Jodie Meeks have all attempted to fill the spot, to varying degrees of success.  Overall, we're mostly talking about small sample sizes.  

After the jump we'll take a look at some statistics from the starting lineups.  All of these statistics do not include last night's loss to the Los Angeles Lakers.

The first graph will take a look at the six different starting lineups used, and the per-game averages the Sixers accrued during those starts.  One key point, this isn't data from just when that lineup was on the court, this was from the entire game.  Columns are the # of games the unit started, the record during those games, average points scored by the Sixers, average points scored by the opponent, and the differential, respectively.

Nothing overly surprising, if you've been following the team.  The Sixers have had the most team success during the string of games since Meeks has been inserted into the starting lineup.  They got blasted in Toronto, the game Brand missed, and started off the year with two tough games against the Heat and Atlanta, two games where despite having the second worst point differential weren't incredibly poorly played games.

The second graph is the same setup, entire game stats from the games each starting lineup was used, not just rotational, but this time with the shooting stats.

Again, taking away the first two games of the year and the Toronto game, the team has been fairly consistent defensively, with big booms in three point efficiency when Kapono or Meeks has been starting.

Same breakdown (per-game, not rotations only), but with ballhandling stats.

What's jumped out to me has been the huge drop in turnovers since Meeks has come in.  This is probably slightly inflated by the most recent game against the Clippers when the Sixers only had 3 turnovers, but they've been doing a good job overall.  The floor spacing Meeks provides probably plays more than a small role in this.

The next graph is the advanced stats from the games each unit has started.  Again, these are for the entire games, not just the rotations of the starting lineup.

Two things really jump out at me:

1) The great play on both sides of the ball for the current starting lineup

2) How much the team struggled defensively when the Holiday/Turner/Iguodala starting lineup was used, and how much success it had offensively.

The next graph is advanced stats as well, but in terms of rebounding, free throws, and turnovers.

The rebounding numbers probably aren't all that relevant, since we're mostly talking about changing perimeter players.  The fairly good job on the defensive glass in the last 7 games probably has more to do with the improved play of Hawes in that regard than anything Jodie Meeks has done.  That being said, it's interesting that the team didn't do better in the games the Turner/Iguodala backcourt started.  Again, the turnover percentage during the last 7 games has been absurdly good.

The next graph is actual rotational data, grabbed only from the time each lineup was on the court.  Note this isn't only from games that they started, although a bulk of the minutes do come from them.

First, it can't be emphasized enough how small of sample sizes these are.  Because of how frequently the starting lineup has been changed, none of them have been on the court very long.  The statistical relevance of 15 minutes of basketball is virtually nothing.  That being said, I think something can be gleaned from the 4 lineups that have played 50+ minutes.  No conclusions, certainly, but hypothesis can be drafted at the very least.

Sadly, only two of the 6 starting lineups have a positive +/-, the one where Turner replaced Iguodala in the starting lineup and the current starting lineup.  Obviously, opponents have played a huge part in this.  Obviously, we've talked about the generally weak competition that the current lineup has gone up against, and the Holiday, Turner, Nocioni lineup played a lot of minutes against Toronto, Cleveland, and an Oklahoma City team that was struggling at the time.  That being said, that lineup did go up and play well against the Knicks and defensively dominated the Bucks.

The next graph is again specific to the rotations time while on the court, this time just normalized out to per-48 minute averages to make it easier to compare, and also the won/loss record of the rotation (meaning, how many times the Sixers outscored their opponents during the time that lineup was on the court, and how many times the Sixers were outscored.  Ties go to the Sixers.

Again, the astounding thing is how infrequently some of the lineups have been on the court together.  It's hard to reall gain much knowledge from 3 trips on the court.

But, again, the production of the current starting lineup is absurd, outscoring the opponent 107.2-84 over a 48 minute period.  

What's also interesting is how much trouble the Sixers had scoring when Kapono and Nocioni were in the lineup, at least until Turner replaced Iguodala.  

Here are some other statistical oddities:

 

• The Sixers have 7 guys averaging in double figures scoring in games the most recent starting lineup has started, with nobody scoring more than 14.6 points.  Jrue Holiday leads the way (14.6), followed by Elton Brand (14.4), Jodie Meeks (14.1), Andre Iguodala (13.7), Thaddeus Young (12.6), Louis Williams (11.4), Spencer Hawes (10.9).
• On a per 40 minute basis, those seven are contributing: Louis Williams (22.5), Thaddeus Young (20.6), Jodie Meeks (17.6), Spencer Hawes (16.6), Jrue Holiday (16.3), Elton Brand (16.2), and Andre Iguodala (14.3).
• Evan Turner has fallen virtually out of the rotation since being removed from the starting lineup, averaging just 13 minutes per game, not counting last night's DNP-CD.  Similar fates have been suffered by Marreese Speights (9:55 per game), Tony Battie (9:49), Andres Nocioni (8:01), and Jason Kapono, who has played 12 seconds since the lineup change.
• The point differential has come almost entirely from the starters during the past 7 games.  Iguodala (+82), Meeks (+80), and Hawes (+79) lead the way, with Holiday (+54), Brand (+50) not far behind.  The highest reserve has been young at +23, no other reserve is in double figures in the positive on +/-.  
• Over the 5 game stretch without Iguodala, Evan Turner averaged 37 minutes, scored 12.8 points on 45% from the field, grabbing 8 rebounds, with 2.8 assists and got to the line 4 times per game.  Turner has averaged 4.9 points in 20 minutes, shooting 40% from the field, grabbing 3.7 rebounds, dishing 1.6 assists and getting to the line only 1.5 times per game for the rest of the season.  Over the last 7 games, Turner has averaged 3.3 points in just under 14 minutes.
• Iguodala was the main benefactor during the first two games Kapono started, averaging 18.5 points, 6.5 rebounds, 7.5 assists and 2 steals on 48.4% from the field during those two games.  He also struggled mightily when Kapono was replaced by Nocioni, averaging only 10.5 points, 4.3 rebounds and 5 assists on 40% from the field.
• Jrue Holiday, on the flip side, struggled the first two games with Kapono (7 points,3 rebounds, 3.5 assists, 25% from the field), but really turned it on after the change with Nocioni being inserted.  He averaged 16.5 points on 52% shooting, to go along with 7 assists and 2.3 steals when he started with Iguodala and Nocioni, then averaged 14.8 points and 8.2 assists when starting with Turner and Nocioni.
• When Iguodala's assist numbers go up, Holiday's go down.  During the first two games Iguodala averaged 7.5 assists per game and Jrue only 3.5.  During the last 7 games starting with Meeks, Iguodala has averaged 6.7 and Holiday's fell to 5.6.  When Iguodala was out, Holiday averaged 8 per game.