Maybe Flip Saunders is a Good Coach

Posted on June 5, 2008 by


The Detroit Pistons won 53 games in 2006-07.  Despite having the best mark in the Eastern Conference, the Pistons were eliminated in the Eastern Conference Finals.

And so the Pistons made changes. 

The Roster Changes of 2007

The following players appeared in a Detroit uniform in 2006-07 but did not suit up for the Pistons in 2006-07:

Carlos Delfino: 1,372 minutes, 0.188 WP48, 5.4 Wins Produced

Chris Webber: 1,277 minutes, 0.124 WP48, 3.3 Wins Produced

Dale Davis: 464 minutes, 0.169 WP48, 1.6 Wins Produced

Will Blalock: 166 minutes, -0.064 WP48, -0.2 Wins Produced

Combined these four players produced 10.1 wins in 3,279 minutes.

Everyone else on the 2006-07 roster returned for at least part of the 2007-08 campaign.  Joining the returning players were eight players.  Four of these players – Jarvis Hayes, Juan Dixon, Walter Herrmann, and Primoz Brezec – played in 2006-07; so we can look at their numbers from that campaign to calculate what the Pistons should have expected in 2007-08.  Theo Ratliff didn’t play much in 2006-07, but we can look at his WP48 [Wins Produced per 48 minutes] number from 2005-06.  For the rookies — Rodney Stuckey, Aaron Afflalo, and Cheikh Samb – we will just take their 2007-08 numbers as given.  With numbers in hand, how many wins could the Pistons have expected from the 2007-08 newcomers?  The answer is reported below.

Jarvis Hayes: 1,287 minutes, -0.029 WP48, -0.8 Wins Produced

Rodney Stuckey: 1,081 minutes, 0.069 WP48, 1.5 Wins Produced

Arron Afflalo: 970 minutes, 0.079 WP48, 1.6 Wins Produced

Juan Dixon: 244 minutes, -0.028 WP48, -0.1 Wins Produced

Theo Ratliff: 223 minutes, 0.065 WP48, 0.3 Wins Produced

Walter Herrmann: 199 minutes, 0.060 WP48, 0.2 Wins Produced

Primoz Brezec: 98 minutes, -0.125 WP48, -0.3 Wins Produced

Cheikh Samb: 31 minutes, 0.058 WP48, 0.0 Wins Produced

Combined these eight players played 4,133 minutes in 2007-08.  Given past performance, these players should have produced 2.6 wins in this amount of playing time. 

So the Pistons lost 10.1 victories and replaced these with players who should have combined to produce less than three.  Given this expectations, we should have expected the Pistons to get worse.

Is it the Coaching?

But this didn’t happen.  The Pistons actually got better.  The team’s victory total increased to 59. Detroit’s efficiency differential – offensive efficiency minus defensive efficiency – increased from 4.59 to 8.17 (the 2007-08 mark was the best in team history).  So how did the Pistons improve?

From efficiency differential we get Wins Produced.  And when we look at Wins Produced we see where the Pistons got better.

Table One: The Detroit Pistons in 2007-08

Looking at Table One we see the most movement in the production of Jason Maxiell.  Had Maxiell maintained his 2006-07 WP48 he would have only produced -0.6 wins in the time he received in 2007-08.  Maxiell’s WP48, though, improved to 0.141. Consequently his Wins Production increased to 5.2.

Although Maxiell’s improvement was impressive (and not entirely surprising in a young player), his leap only explains about a third of the additional wins we saw from the Pistons.  For most of the additional wins not linked to Maxiell we have to turn to Detroit’s veterans.  Chauncey Billups, Rasheed Wallace, Richard Hamilton, and Tayshaun Prince  – yes, the veterans who supposedly didn’t respect Flip Saunders – produced 10.1 more wins than their past performance would suggest.

Despite these additional wins, though, the Pistons still faltered in the Eastern Conference Finals.  As a consequence, Flip Saunders is now unemployed.

Let’s reveal, though, what we know about the Pistons:

1. Given the past performance of the players the Pistons employed in 2007-08, the Pistons should not have been able to secure home court advantage in even the first round of the NBA playoffs. 

2. Although the Pistons improved and posted a franchise best efficiency differential, the Boston Celtics – who eliminated the Pistons in the Eastern Conference finals – posted a better differential (10.95).  In other words, we should not have expected the Pistons to advance to the NBA Finals.

Given these two observations, it’s not clear that Saunders should have been let go.  Yes, people say the Pistons were not listening to Saunders.  Whether this is true or not, we do see evidence that several individuals improved under their coach.

Now can we attribute this improvement to Saunders?  That’s not entirely clear.  Player performance can change because of experience, injury, roster instability, changes in minutes played, and diminishing returns.  Even when you control for all these factors (which Mike Leeds, Eva Leeds, and Mike Mondello have done in a working paper), though, some coaches (not all) appear to have an impact (in other word, it is possible some coaches make players better).  And one of the coaches filed under “some” is Flip Saunders (in other words, it is possible – not a certainty or the “truth” – that Saunders makes players better).  So perhaps letting Flip go is not the best move Detroit could make.

Some Other Moves

Regardless of what the data says, though, Flip is now gone.  And changing coaches is not the only move the Pistons might make.  There are rumors that Rasheed Wallace is also on the way out.  If moving ‘Sheed clears the way for Amir Johnson, then this move should help.  In 978 career minutes, Amir Johnson has posted a 0.238 WP48.  The best mark Rasheed ever posted in his career was 0.142, which he posted in 2007-08, 2001-02, and 1996-97 (the mark in 2001-02 was actually 0.143).  If Amir can maintain his production – and he actually gets to play most of Rasheed’s minutes – the Pistons are better off moving Rasheed elsewhere. 

In addition to switching R. Wallace and A. Johnson, the Pistons might benefit from another change.  And that would be a change in expectations. 

For the past seven seasons the Pistons have averaged 55 regular season wins per year.  No team in the Eastern Conference can best this mark (and only Dallas and San Antonio have done better in the West). Despite this record, though, Detroit has only won one title and advanced to the NBA Finals twice. This lack of post-season success – amidst all this regular season accomplishment – has caused fans of the Pistons much anguish. 

Perhaps an attitude adjustment, though, would be helpful.  Only one team will win a title each year.  Fourteen NBA teams don’t even make the playoffs and of those that enter the post-season, fifteen have to finish with a loss.  So if you are only happy if your team wins a title, then you are going to be disappointed almost all the time.

Remember, almost half of all NBA franchise have yet to win a title.  For example, the Dallas Mavericks have been even better in the regular season the past seven years and they still haven’t won a championship.  If every time you lose in the playoffs you fire your coach and/or change key players – as some fans of the Pistons seem to suggest – then you might be changing coaches and your roster constantly.  And although we are not certain about what causes a player to be better, we do have evidence that moves like changing the coach or dramatically re-making the team’s roster – by themselves – make players perform worse.

In sum, perhaps fans of the Pistons should be a bit happier with this team’s amazing run.  The Detroit franchise has never posted 50 or more wins in seven consecutive seasons.  It’s an amazing accomplishment, achieved entirely without a single player most NBA observers would label of “superstar.” Such an accomplishment should be applauded and appreciated.

Or if you don’t wish to applaud, you can ignore my advice and just get mad because the Pistons had the temerity to lose to a better team.  But if that’s going to be your attitude, be prepared to be mad most of your life.

– DJ

The WoW Journal Comments Policy

Our research on the NBA was summarized HERE.

The Technical Notes at provides substantially more information on the published research behind Wins Produced and Win Score

Wins Produced, Win Score, and PAWSmin are also discussed in the following posts:

Simple Models of Player Performance

Wins Produced vs. Win Score

What Wins Produced Says and What It Does Not Say

Introducing PAWSmin — and a Defense of Box Score Statistics

Finally, A Guide to Evaluating Models contains useful hints on how to interpret and evaluate statistical models.