NBA.com writer John Schuhmann posted an article on the NBA's Hangtime Blog laying out a statistical case for George Karl of the Denver Nuggets to win the Coach of the Year award yesterday, but I think his data mining came up with more lead than solid gold.
Schuhmann's case was based on the fact that Karl transformed the Nuggets from the most efficient offense in the NBA before trading Carmelo Anthony to the league's most efficient defense after the trade. While Schuhmann concedes that the shift in Denver's offensive and defensive stats have something to do with the players, I don't think he gave them enough credit.
NBA writers and analysts tend to give coaches credit for surprising outcomes that can't be explained by scoring because defensive statistics are not as mature as offensive statistics. Dan LeBatard of the Miami Herald illustrated this point in his column about people voting Derrick Rose for MVP because they don't understand that the Chicago Bulls defense is why they have the best record in the NBA Eastern Conference.
The Nuggets defense has improved because several players have improved their defensive contributions after the Melo trade and not because Karl has become a better defensive coach after the All-Star break. One of the best ways to illustrate this is with the Win Score and Wins Produced statistical models developed by sports economist David Berri from the Wages of Wins Journal.
The Win Score and Wins Produced models measure how much a player's box score statistics contributed to their team's efficiency differential and wins. An average player produces an estimated 0.100 wins per 48 minutes (Est.WP48), a star player produces 0.200+ Est.WP48 and a superstar produces 0.300+ Est.WP48. More information on these stats can be found at the following links:
This article will use those models to analyze the following:
- How the Nuggets defense has changed,
- How the Nuggets rotation has changed since the trade and
- How the Nuggets new rotation has changed their defense.
The Nuggets have made the following improvements on defense since the Melo trade:
Shooting: Opponents' shooting efficiency has decreased by 2.6 percent and the Nuggets' blocks have increased by 1.6 per game.
Protecting the glass: Defensive rebounding has increased by 2.2 per game, but opponents are only missing an average of one more shot per game. The Nuggets have increased their defensive rebounds by decreasing their opponents' offensive rebounds by 1.8 per game.
Pressuring the ball: Nuggets have allowed 2.4 less assists, forced 1.2 more turnovers and grabbed 1.6 more steals per game.
Fouling: Nuggets committed 1.7 less fouls per game and allowed 3.1 less free-throw attempts per game.
Scoring: Lower shooting percentages, more turnovers and less free throws have resulted in 8.1 less points allowed per game.
With the Nuggets defensive improvements after the trade was established, the next step is to identify how the trade changed the Nuggets rotation and whether those changes improved the defense.
Based on lineup data from basketballvalue.com, the big changes in the Nuggets rotation have been at point guard and small forward.
At point guard, the combo of Chauncey Billups and Ty Lawson was replaced by Ty Lawson and Raymond Felton after the trade.
At small forward, minutes that were used by Melo and J.R. Smith have been replaced by Wilson Chandler and Danilo Gallinari although Smith did play a lot of small forward while Gallinari was injured.
At shooting guard, the rotation is still Arron Afflalo and J.R. Smith, at power forward the rotation is still Kenyon Martin and Al Harrington and at center the rotation is still Nene and Chris "Birdman" Andersen.
As this spreadsheet illustrates, the new combinations at point guard and small forward had worse defensive ratings than the combinations Coach Karl used before the trade and that leads to a few questions:
- Did these players improve their defensive stats after the trade?
- If these players did improve their defense, then was it a result of Karl "coaching them up" or was it something they had done before in their careers?
What impact has Raymond Felton had on the Nuggets defense?
The new point guard rotation of Lawson and Felton had a positive effect on the Nuggets defense in the following areas (as illustrated in this spreadsheet):
Shooting: Point guards have shot one percent worse vs. Lawson-Felton than they did vs. Billups-Lawson even though the new rotation has blocked less shots.
Protecting the glass: The Lawson-Felton rotation has averaged 2.1 more defensive rebounds per 48 minutes than Billups-Lawson did. Felton has increased his defensive rebounds by 1.8 per 48 minutes since the trade.
Pressuring the ball: Since Felton averaged more steals than Billups, this area was expected to improve after the trade. The Lawson-Felton rotation has averaged 1.4 more steals per 48 minutes than Billups-Lawson, but the biggest change has come from Lawson increasing his defense by one more steal per 48 minutes. Point guards have also averaged 0.7 less assists per 48 minutes vs. Lawson-Felton than they did vs. Billups-Lawson.
Fouling: Since Felton averaged fewer fouls than Billups, this was area was also expected to improve after the trade, but it has improved more than expected. The Lawson-Felton rotation has averaged 0.9 less fouls per 48 minutes than Billups-Lawson and as a result, point guards are getting one fewer free throw attempt per 48 minutes vs the Nuggets since the trade.
Scoring: With a lower shooting percentage and fewer free throw attempts, point guards have scored 2.7 less points per 48 minutes vs. the Lawson-Felton rotation than they did vs. Billups-Lawson.
If they had maintained the same level of production they showed before the trade, then the combination of Lawson and Felton at point guard would have been slightly less productive than the combination of Billups and Lawson. The improvements on defense, however, have resulted in a 16 percent increase in their production and a seven percent decrease in the opponent's production.
The defensive improvements by Lawson and Felton translate to an extra win for the Nuggets over 82 games.
The new small forward rotation of Chandler and Gallinari has affected the Nuggets defense in the following areas (as illustrated in this spreadsheet):
Shooting: Small forwards have shot three percent worse vs. the Nuggets since the trade. The Chandler-Gallinari rotation averaged more blocks than Anthony-Smith did before the trade and continued to do so after the trade but Gallinari has unexpectedly increased his blocks by 0.3 per 48 minutes.
Protecting the glass: Chandler and Gallinari were worse rebounders than Anthony and Smith before the trade with 3.4 less defensive rebounds per 48 minutes. After the trade, however, they have improved to close that deficit down to just 0.6 less defensive rebounds per 48 minutes. Gallinari has increased his defensive rebounds by 2.3 per 48 minutes.
Pressuring the ball: Small forwards have averaged 0.6 more turnovers per 48 minutes vs. the Nuggets since the trade. Chandler and Gallinari have increased their steals by 0.8 per 48 minutes to equal or surpass Anthony in that category, but trail Smith (who played a significant number of minutes at small forward while Gallinari was injured).
Fouling: Chandler and Gallinari averaged less fouls than Anthony and Smith before the trade, but they have decreased their fouls even further and average 1.8 less fouls per 48 minutes since joining the Nuggets. As a result, small forwards have averaged 1.1 fewer free throw attempts per 48 minutes vs. Denver since the trade.
Scoring: With decreased shooting efficiency, fewer trips to the line and more turnovers, small forwards have scored 1.3 less points per 48 minutes vs. the Nuggets since the trade.
Across offensive and defensive stats in the box score, Chandler and Gallinari's combined production has been the same as it was before the trade (Chandler's has decreased some while Gallinari's has improved by nearly the same amount).
That level of production is seven percent less than what Anthony and Smith provided at small forward before the trade, but the defensive improvements at the position since the trade have decreased the opponents' production by 33 percent. Additionally, without those defensive improvements, Chandler and Gallinari would be producing 31 percent less than Anthony and Smith did before the trade.
The defensive improvements at small forward translate to an extra two to three wins for the Nuggets over 82 games.
George Karl and Nuggets fans are waiting to see what happens with the team moving forward.
The changes Karl made in the Nuggets rotation after the trade has resulted in players getting minutes that improved the team's defensive stats, and this improvement is worth at least three or four extra wins over an entire season.
The final question is, can this improvement be attributed to Karl's coaching or the players themselves?
As the head coach, Karl obviously determines the rotation but this rotation was mostly pre-determined for him. Felton had to get minutes at point guard after the trade because both of the players that received minutes at the position before the trade (Billups and Anthony Carter) were no longer on the roster.
The same goes for Chandler and Gallinari at small forward. The choice was to either fill the void left by Anthony at small forward with rookie Gary Forbes or give the minutes to the new players joining the roster.
Did Karl "coach up" the players in his new rotation to improve their defensive numbers?
Felton's rebounding numbers are 1.7 boards higher than the per 48 minutes average for his career, but the rest of his numbers are similar to his career averages. Gallinari's rebounding numbers have seen the same improvement as Felton's in terms of his career averages, and his fouls have improved by 0.9 more than his career numbers so far, but his steals and blocks are similar to what he's done previously.
Since the conventional wisdom is that rebounding and defense are about effort, it would be counter-intuitive to give Karl credit for the players' performance in those areas unless he's such a good motivator that he inspires the players to increase their effort beyond what they normally provide on a consistent basis. That seems like a flimsy argument to support a nomination for the Coach of the Year award.
Maybe the players have performed better because they are just happier, like the Nuggets' former Director of Quantitative Analysis suggested on the ESPN Truehoop blog.
That would explain more of the team's improved performance than the defensive improvements at point guard and small forward, which only account for three to four wins in what would amount to a 24-win improvement over the course of a season if the Nuggets could continue to play the way they have since the trade.
That suggests the Nuggets who were already on the roster are more responsible for the team's improvement than the ones they got from the Knicks. If they can be this productive after the trade, then the question is, why didn't Karl get them to be this productive before the trade?
The answer to that question would probably explain why Karl isn't the Coach of the Year. Because if those players revert back to their previous levels of performance, then the Nuggets won't win as many games as they have since the trade despite Karl continuing to coach them the same way.