Yes, the Knicks have Improved

Posted on March 26, 2009 by


A few days ago the following article was posted by Mike Kurylo of Knickerblogger:

Haven’t the Knicks Improved?

Kurylo’s column was in response to an article in Pro Basketball News – Courtside View: Knicks said to be improving, but are they really? – that argued the Knicks have not really gotten better in Mike D’Antoni’s first season as head coach.  The following paragraph from this article specifically attracted Kurylo’s attention.

The Knicks have improved from 21st in scoring last season (96.9 ppg) to fourth this season (105.6 ppg) but they have dropped from 22nd in points allowed (103.5 ppg) to 28th (108.2 ppg) and their point differential of -2.6 ppg ranks in the bottom third of the league (22nd), only a few spots better than last season (-6.6 ppg, 25th). The Knicks were last in field-goal percentage differential last season (-.036) and they are last again this season (-.038). Although Lee has emerged as a nightly double-double threat, the Knicks have markedly declined overall on the boards, dropping from 18th in rebounding differential (-.1) to 27th (-3.9). A team that consistently gets out-shot and out-rebounded obviously has no realistic chance to be successful, no matter how many points it scores or how many players post career high individual numbers.

The primary issue Kurylo raised upon presenting this paragraph is that the author focused on points-per-game, rather than points-per-possession.  The former ignores pace and consequently it doesn’t accurately reflect the quality of a team.  Having read both the article in Pro Basketball News and Kurylo’s discussion, I thought I would spend today’s column reflecting on the dispute.

The Knicks Have Improved

Let me begin with another quote from the Pro Basketball News article:

“….the Knicks are not much better than they were when Isiah Thomas ran the show; Thomas had a .402 winning percentage in his first season on New York’s bench, virtually identical to the .406 winning percentage that D’Antoni has posted so far in his first season in New York.”

As Kurylo notes, such a comparison is cherry-picking at its worst.  D’Antoni did not inherit the Knicks from 2004-05.  The team he inherited from Isiah Thomas posted a 23-59 record last season.

And if we compare the 2007-08 team to the Knicks in 2008-09, it’s clear this team has improved.

In 2007-08 the Knicks scored 101.2 points per 100 offensive possessions and allowed 108.1 points per 100 defensive possessions.  So the Knicks efficiency differential – offensive efficiency minus defensive efficiency – was -6.9.  Obviously this is bad.  In fact, it was the worst differential for the Knicks since these numbers could be fully calculated (complete calculations go back to 1973-74).  

Now that we have established what D’Antoni inherited, let’s make a proper comparison.  After 71 games in 2008-09 the Knicks have posted the following numbers:

Offensive Efficiency: 105.2 points per 100 possessions

Defensive Efficiency: 107.8 points per 100 possessions

Efficiency Differential: -2.6

Looking at these numbers it’s clear the Knicks have improved on offense and defense.  Consequently it’s not surprising that this team has already surpassed last year’s wins.

Why Have the Knicks Improved?

Much of what I said was already noted in Kurylo’s column.  Now let me offer something more.

We can see in the efficiency differential numbers that the Knicks are better.  What we want to know, though, is why.  And to answer this question, we should look at the performance of the individual players.

As I have noted in the past, Wins Produced translates the efficiency differential story from the team to the players.  And when we look at the players currently on the Knicks via Wins Produced, one can see who is responsible for this team’s improvement.

Table One: The New York Knicks after 71 games in 2008-09

Table One presents the standard views offered in this forum.  First we have how many wins the Knicks could expect if we made the assumption that what we saw last year from these players would be seen in 2008-09 (by the way — as I noted a few days ago –this exercise should not be thought of as a complete forecast).  This initial view suggests that Knicks should have already won 24 games this year.

We then have a second view which looks at how many wins the Knicks should expect given the performance of these players in 2008-09.  This view suggests the Knicks should have 29 victories after 71 games this season.

If we put both views together we can see that the Knicks might have improved because a) the team has changed its roster; and b) some players on this roster have improved.

            The Roster Moves

The many roster moves the Knicks have made began in the off-season and continued into the regular season.  Looking at these moves, here are the players who played for the Knicks last year who did not appear this season (or in the case of Eddy Curry not very much): Renaldo Balkman, Mardy Collins, Eddy Curry, Jamal Crawford, Jerome James, Fred Jones, Stephon Marbury, Randolph Morris, Zach Randolph, and Malik Rose.  These players played 11,792 minutes last year and produced 8.8 wins.  These numbers give us a 0.036 WP48 [Wins Produced per 48 minutes].  Average is 0.100, so these players were collectively below par (although Balkman and Randolph were above average).   

The players added to this roster – who are still on the team – include the following:  Chris Duhon, Danilo Gallinari, Al Harrington, Larry Hughes, and Chris Wilcox.  These five have played 5,629 minutes and produced 6.0 wins for a WP48 of 0.051.  When we look at the performance of the specific players, only Duhon [2,563 minutes, 6.2 Wins Produced, 0.115 WP48] has been above average (and as noted last September, Duhon’s performance is not surprising).  Despite the addition of Duhon, the roster moves the Knicks have made have not made much difference.  The players added are not much different from the players lost.

One should add, though, that the loss of Curry (at least he is lost to the team on the court), Crawford, and Marbury should have harmed the Knicks.  At least, if you believe scoring matters, the loss of these scorers should have cost the Knicks in the win-loss columns.  But that is not what has happened (and no, it’s not because the team added Harrington and Hughes).

            Improving Players

If the improvement is not primarily from the roster moves then we have to look at the performance of the individual players who remained.  The Knicks only have the following five players from last year’s team still on the roster (and actually playing): Wilson Chandler, David Lee, Jared Jeffries, Quentin Richardson, and Nate Robinson.  These five posted the following numbers last year: 8,090 minutes played, 14.7 Wins Produced, and 0.087 WP48.  This season, after 71 games, this quintet has played 9,467 minutes, produced 23.6 Wins Produced, and posted a 0.119 WP48.  When we look at the individual players we see that Wilson Chandler is actually offering a bit less.  And Jeffries is about the same.  But Lee, Richardson, and Robinson are doing more.

Of these, the biggest improvement in per-minute production is seen in the play of Nate Robinson.  And as Table Two indicates, this improvement can be tied to increased production with respect to shooting efficiency, rebounds, steals, and assists.   In sum, Robinson has gotten better in a number of areas.

Table Two: Evaluating Nate Robinson

Like Robinson, Lee’s WP48 is higher this year than it was last year (although not what it was two years ago).  Lee’s increase in Wins Produced, though, is nearly as big as the increase we see from Robinson.  Lee produced 12.7 wins last year in 2,356. This year, assuming he continues to play 35 minutes a game, Lee will produce 17.1 wins in 2,864 minutes.  So Lee will offer more than four additional wins, and much of this addition is simply due to the fact that Lee is finally getting more than thirty minutes of playing time per game. 

Putting the Picture Together and Looking to the Future

Let’s go back to the beginning.  Projecting to the end of the 2008-09 season we see that the Knicks will improve by about 10 wins this year.  Despite numerous roster moves, it’s not the players coming and going that have really produced this change.  Most of the change in wins can simply be tied to the improved play of Robinson (which happens with young players) and the increased playing time for David Lee.

These two players, along with Duhon, produce almost all of the wins on the Knicks.  This trio – who are scheduled to make about $10.5 million next season (after making $9.2 million this year) — are the only above average players currently on the roster.  Unfortunately the Knicks are scheduled to pay $60 million more to a collection of players who really have not helped this year.  So without further roster changes it seems unlikely the Knicks are going to improve much more.

Of course, given what we have seen over the past few months, further roster changes are quite likely. One suspects that this team will be somewhat different at the start of the 2009-10 season. And it’s well-positioned (as are a number of other teams) for the 2010 free agent market.  So the improvement we have seen this year – and yes, this team is definitely better – can continue into the future.  That is if this team is able to find additional players who can produce wins. 

– 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.