Saturday, January 8, 2011

Heat Check: Benching John Hollinger

Last month, at the request of reader Giovanni Valladares, I took a look at where the Miami Heat's bench ranked amongst the rest of the reserves in the NBA.  This month, John Hollinger did the same for espn.com in NBA: Rating the top benches - ESPN (insider access required).

In the piece I posted last month, the Miami Heat had the second-most productive bench in the NBA, despite the loss of Udonis Haslem and Mike Miller to injury. Hollinger, however, reports that the Heat have the second-worst bench in the league. What gives?

The big difference is that Hollinger used his Player Efficiency Rating (PER) and I used Wins Produced. For the differences between PER and Wins Produced, see The Wages of Wins Journal FAQ.



This article will use Win Score and Wins Produced, statistical models created by Professor David Berri from the Wages of Wins Journal, to measure how much a player's box score statistics contributed to their team's performance. An average player produces 0.100 wins per 48 minutes (WP48), a star player produces 0.200+ WP48 and a superstar produces 0.300+ WP48. More information on these stats can be found at the following links:


Ranking Reserves By Hollinger's Bench Rating vs. WP48
The top five benches according to Hollinger's rating were:
  1. Toronto Raptors
  2. Denver Nuggets
  3. Philadelphia 76ers
  4. San Antonio Spurs
  5. Houston Rockets
Last month, the top five benches according to Wins Produced per 48 minutes (WP48) were:
  1. Orlando Magic
  2. Miami Heat
  3. Los Angeles Lakers
  4. Dallas Mavericks
  5. Oklahoma City Thunder
There are no similarities at the top of the lists. What about the bottom?

The bottom five benches according to Hollinger's rating were:
  1. New Orleans Hornets
  2. Miami Heat
  3. Los Angeles Clippers
  4. Minnesota Timberwolves
  5. New Jersey Nets
The bottom five benches according to WP48 were:
  1. Golden State Warriors
  2. New York Knicks
  3. Minnesota Timberwolves
  4. Los Angeles Clippers
  5. Memphis Grizzlies
There are two teams in common at the bottom of the lists - the Timberwolves and the Clippers. Let's see if updating the WP48 rankings for NBA benches finds more common ground. The method I used to calculate the WP48 for each team's bench was simple:

  1. Identify each team's starting lineup at basketball-reference.com
  2. Sum the wins produced for the non-starters using data from NerdNumbers (or from the Heat Produced page for Miami)
  3. Divide the sum of the reserves' wins produced by the minutes played
  4. Multiply the result by 48.


With the updated numbers, there is still no common ground in the top five but both rankings list the Lakers, Magic, Raptors, Spurs, Nuggets and 76ers in the top 10. None of those teams have the same rankings on both lists. Both rankings also have the Clippers, Hornets and Timberwolves in the bottom five and the Trailblazers, Celtics and Warriors in the bottom 10. The Blazers and Warriors have the same ranking on both lists.

But enough of Hollinger's bench rankings, they're not that important because they're based on a model that doesn't explain wins very well and that's what the Miami Heat Index is about - wins produced.


Heat's Declining Bench Production
Let's look at changes in the top five benches over the last month, according to WP48.

The Heat's bench production has fallen from second in the NBA for games played through December 8th to fifth in the NBA for games played through January 7th. What's happened in the last month?



As you can see in the spreadsheet above, the bench's per-minute production has declined by 28 percent in the last 29 days. The biggest declines have come from James Jones and Juwan Howard. Those two combined to produce an estimated 0.9 wins less than expected in the last month. The spreadsheet below shows the changes in their production over the last month.



Let's start with Juwan Howard. In my last look at the bench, I said "As for Juwan Howard, he's been bad at everything except shooting and passing the basketball..."

Well, since then Juwan has managed to become a bad shooter and below average passer. He did improve his rebounding and got to the line a little more (even though he's still below average at the position) and decided to stop beating the living hell out of opponents (his fouls declined from 5.9 per 48 minutes to 3.3 per 48 minutes). 

Those small improvements weren't enough to make up for his terrible shooting over the last month. Juwan's adjusted shooting percentage for the last 14 games is just 26 percent. If Mike Miller had managed to make just one more shot, he would be shooting better than Juwan Howard in the last month and Miller's only made one shot since he returned from injury. 

While I'm on the topic of Miller, the bench would be much more productive if he wasn't shooting worse than Juwan Howard. Miller's performance has been the third-most disappointing in the last month with an estimated WP48 of -0.280. Last season, Miller produced 0.243 WP48 for the Washington Wizards. If he played at that level of production in December, then all of the bench's decline would have been wiped out.

That's enough of the players that can't shoot. Let's move on to the Heat's designated shooter, James Jones.

Jones' performance declined in terms of shooting efficiency and volume, scoring, turnovers and fouls, but overall his performance remained the same with an 8.2 Win Score. The problem for Jones was that the average performances at the positions he plays increased.

The average Win Score per 48 minutes was 6.1 for a shooting guard and 7.8 for a small forward in the first six weeks of the Heat season and Jones' production was above average with an estimated 0.133 WP48. Since then, the average Win Score per 48 minutes has increased to 9.0 for a shooting guard and 10.1 for a small forward and Jones' production (which essentially stayed the same) is now below average with an estimated 0.044 WP48.

You can probably guess the big reasons for this shift in the average production at shooting guard and small forward in Heat games - their names are Dwyane Wade and LeBron James aka Eastern Conference Players of the Month. In the first six weeks of the season, Wade produced a 10.4 Win Score per 48 minutes while LeBron produced an 11.4 Win Score. In the last month, Wade has increased his Win Score to 15.5 and LeBron has increased his to 16.

The result is that the average production at those positions has increased by 37 percent and it's gotten to the point where having Jones on the court, instead of Wade or LeBron, is a negative value proposition for the Heat.

I think that's a good thing.

The Heat have not been affected by the bench's decline because the starters have increased their per-minute production by 31 percent. Since they play so many more minutes, it's much more important for the starters to produce than the reserves. Wade and LeBron are now playing like superstars with each averaging more than an estimated 0.300 WP48 since December 9th. Bosh is averaging an estimated 0.248 WP48 since then.

According to NerdNumbers, the Celtics are the only team with two starters (Rajon Rondo and Kevin Garnett) averaging 0.300+ WP48 and a third starter (Paul Pierce) averaging 0.200+ WP48. So if the Heat keep playing the way they have been since December 9th, they'll be in good shape.

Garnett's production will likely decline with injuries and Pierce probably won't sustain his current level of production, either (he finished last season averaging 0.163 WP48 - shoutout to NerdNumbers). The bench is already better than the Celtics' declining bench and it will only improve as Miller improves. Miller is important because he can remain a productive player for the Heat even as LeBron and Wade raise the bar with their play while Jones may not be able to.

Another good thing for the Heat is the improved play of the centers. Despite all of the criticism of the Heat having no reliable big men, they've performed quite well in the last month:
  • Zydrunas Ilgauskas improved his production from an estimated 0.130 WP48 in the first six weeks of the season to an estimated 0.185 WP48 over the last four weeks.
  • Joel "Doc the Warden" Anthony has improved from an estimated -0.041 WP48 to 0.090 WP48. If he continues to find a way to be productive while locking down the opponent's best power forward, it will make the Heat VERY tough to score against without sacrificing too much production in other areas.
  • Erick Dampier has improved from an estimated  0.074WP48 to an estimated 0.159 WP48. Here's what I said about Dampier in last month's bench review, "If Dampier improves to his projected level of productivity, then the productivity off the bench will get closer to its UD-levels." Consider Dampier improved. If Miller can keep up his end of the deal, then the bench can start being the best in the league.

Back to the Bench Rankings
Now that the Heat's bench performance has been thoroughly broken down, let's take a quick look at some of the significant changes in bench performance over the last month.
  • Last month, benches for the Hawks, Warriors, Clippers, Grizzlies, Timberwolves, Hornets and Knicks were losing games. A month later, the Hawks and Grizzlies are getting positive production from their reserves. Jordan Crawford, Maurice Evans and Josh Powell are the only reserves that aren't playing better in Atlanta while Tony Allen and O.J. Mayo got along well enough to produce an extra 0.8 wins for Memphis off the bench.
  • While the Hawks and Grizzlies escaped the negative zone, the Celtics bench production fell into it by declining from 0.010 WP48 to -0.001 WP48. The declining production of Glen "Big Baby" Davis and Luke Harangody cost the Celtics 0.6 wins produced off the bench in the last four weeks.
  • The Cavaliers joined the Celtics in the negative zone by moving productive reserves like Daniel Gibson (0.112 WP48) and Antawn Jamison (0.026 WP48) into the starting lineup and replacing them with the negative production of Joey Graham (-0.089 WP48) and J.J. Hickson (-0.108 WP48).
  • The Bulls bench jumped four spots from #7 to #3 by improving their production from 0.071 WP48 to 0.093 WP48. Omer Asik, Kurt Thomas and C.J. Watson made the biggest improvements. Asik improved from 0.005 WP48 to 0.111 WP48, Thomas improved from -0.472 WP48 to 0.048 WP48 and Watson improved from -0.055 WP48 to -0.020 WP48.
  • While Bulls moved up the charts, the Mavericks moved down from #4 to #13. Shawn Marion's production declined from 0.173 WP48 to 0.135 WP48 and Jason Terry's production declined from 0.113 WP48 to 0.063 WP48.
  • The Warriors had the worst bench in the league last month. Vladimir Radmanovic improved from -0.078 WP48 to 0.031 WP48 and Louis Amundson improved from -0.180 WP48 to -0.048 WP48 to help them climb out of the basement.
  • The return of Andrew Bynum helped the Lakers go from #3 to #1 by moving Lamar Odom (0.259 WP48) to the bench.
  • The Thunder moved from #5 to #2 thanks to James Harden improving from 0.080 WP48 to 0.166 WP48 and Nick Collison improving from -0.032 WP48 to 0.063 WP48.
  • The Kings decline from #6 with 0.073 WP48 to #15 with 0.046 WP48. Samuel Dalembert's production dropped from 0.153 WP48 to 0.118 WP48 and Pooh Jeter fell off from 0.351 WP48 to 0.045 WP48.
  • The Raptors moved from outside the top 10 at #13 with 0.052 WP48 to #5 with 0.074 WP48. Julian Wright improved from 0.055 WP48 to 0.109 WP48 and Leandro Barbosa improved from -0.043 WP48 to -0.007 WP48.

That's it for now. I'll follow-up on this story as Mike Miller begins to improve and we'll see where the Heat reserves stand then.

All stats for the Miami Heat were taken from the Heat Produced Page. All other stats were Powered By NerdNumbers.

No comments:

Post a Comment

Post a Comment