April 2, 2010
Playing Time
Every Team has a Story

by Bradford Doolittle

You can tell a lot about a team's season simply by the players that it has used. Or, to put it another way, the story of a team's season is told by the players it has had available.

Each summer, as NBA general managers are piecing together their rosters for the upcoming season, people like us painstakingly play along and try to project the results of that roster. It's a tricky business, fraught with uncertainty, and doomed to fail. Yet the process informs us what to expect and our mistakes give us a deeper understanding of the team-building process.

One of the most inexact parts of projecting team performance is trying to estimate how much playing time a given player is going to receive. Health, productivity, team chemistry and coaching whims play havoc with attempts to forecast court time based on past patterns. However, having that baseline of expected court time gives us a chance to evaluate not just the effectiveness of our projections, but also to craft the story of each team's season.

ROTATION MINUTES VS. PROJECTION
TEAM  PROJ    ACT   %PROJ  RANK
GROUP 1
atl   15189  18285  120.4%   1
okc   14907  16638  111.6%   2
den   14379  16025  111.4%   3
bos   14555  16078  110.5%   4
hou   13674  14966  109.5%   5
------------------------------
GROUP 2
phx   14449  15771  109.1%   6
lal   15330  16591  108.2%   7
min   13956  15067  108.0%   8
nwo   14097  14716  104.4%   9
phi   14837  15453  104.2%  10
------------------------------
GROUP 3
dal   15330  15848  103.4%  11
mem   15365  15808  102.9%  12
tor   14731  15071  102.3%  13
chi   14837  15100  101.8%  14
cle   14978  15207  101.5%  15
------------------------------
GROUP 4
mia   15154  15272  100.8%  16
orl   14872  14684   98.7%  17
sas   14167  13912   98.2%  18
lac   14837  14213   95.8%  19
uta   15084  14445   95.8%  20
------------------------------
GROUP 5
ind   13956  13131   94.1%  21
sac   15119  13858   91.7%  22
njn   14696  12631   86.0%  23
nyk   13956  11963   85.7%  24
cha   14273  11884   83.3%  25
------------------------------
GROUP 6
gsw   13533  10910   80.6%  26
mil   14625  11789   80.6%  27
det   14590  11437   78.4%  28
was   15119  11387   75.3%  29
por   15683  11640   74.2%  30

This is a comparison of the projected and actual minutes for each team's eight-man core generated by NBAPET before the season. The actual numbers have been prorated for 82 games. The core was determined by the eight players forecast to compile the most Wins Produced on each teams roster at the time the season tipped off. The actual minutes represent the total court time that projected core has seen.

There are many reasons to explain the differences between projection and reality. Forecast error, injuries, trades, loss of effectiveness, depth chart changes, shortened rotations, taking your guns into the locker room--the story is different for each team, but this is the story we're looking to tell today.

GROUP 1
TEAM  PROJ    ACT   %PROJ  RANK
atl   15189  18285  120.4%   1
okc   14907  16638  111.6%   2
den   14379  16025  111.4%   3
bos   14555  16078  110.5%   4
hou   13674  14966  109.5%   5

A big factor in the Hawks' rise this season has been the extraordinary record of health posted by their core players. Al Horford, Josh Smith and Mike Bibby have all started every game. Marvin Williams has missed one; Joe Johnson has sat twice. Of the top reserves, Jamal Crawford has missed one game, Maurice Evans three and Zaza Pachulia four. It's remarkable. In Oklahoma City's case, starters Kevin Durant, Russell Westbrook, Jeff Green and Thabo Sefalosha have all played in every game, while Nenad Krstic has missed two. That kind of continuity has helped the Thunder's young core coalesce. In addition, a couple of roles have evolved differently than NBAPET forecast. Instead of James Harden dominating the minutes at two, he's missed some time and shared the role with Sefalosha. Also, instead of Sefalosha sharing supporting backcourt roles with Shaun Livingston and Kyle Weaver, he's on pace to play over 2,300 minutes. Livingston was waived early in the season and injuries have limited Weaver to 144 minutes.

In Denver, Nene, who missed 165 games from 2006 to 2008, has played in all 76 of the Nuggets' games and is averaging a career-high 33.8 minutes per contest. For the Celtics, the story has been foreseeable decline. Kevin Garnett and Paul Pierce have missed time, but in both cases, that was projected. NBAPET did not see Ray Allen playing in all but one contest, nor did it see Rajon Rondo's minutes jumping 3.3 minutes per game. (Perhaps it should have.) The projected minutes of Houston's core of Aaron Brooks, Trevor Ariza and Luis Scola all lagged because of evolving usage patterns. These numbers obviously skew any insights into the Rockets' overall team health because we knew before the projection season that Yao Ming would miss the entire campaign.

GROUP 2
TEAM  PROJ    ACT   %PROJ  RANK
phx   14449  15771  109.1%   6
lal   15330  16591  108.2%   7
min   13956  15067  108.0%   8
nwo   14097  14716  104.4%   9
phi   14837  15453  104.2%  10

For Phoenix, the effect caused by Leandro Barbosa's injury was offset by an unforeseen increase in court time for Jared Dudley. Meanwhile, Amare Stoudemire has played in every game after missing 29 games last season. Also, NBAPET was simply wrong in predicting age-related dips in time for Steve Nash and Grant Hill. Kobe Bryant, after a two-year down trend in his average minutes, has suddenly jumped back up to 2.7 minutes per game more than last season for the Lakers, though he's missed five games. Also, NBAPET thought Derek Fisher's minutes would dip severely. They haven't, but perhaps they have.

Minnesota was a hard team to read before the season. There has been a lot of flux in Kurt Rambis' rotation, but in terms of how the Wolves were projected, the net effect is that the players were tabbed as the team's core have seen more time than anticipated, particularly Jonny Flynn. In New Orleans, the minutes lost by Chris Paul when he was hurt have been offset by bulging minutes for Darren Collison. Also, David West's minutes projection was still being weighed down by the 30 games he missed in the 2006-07 season, which suggests a needed tweak to the NBAPET machinery. For Philadelphia, it'd have been hard to project the 3,200 minutes Andre Igoudala is on pace for. Also, Elton Brand has been more durable than predicted and Jrue Holiday has been used more than anticipated.

GROUP 3
TEAM  PROJ    ACT   %PROJ  RANK
dal   15330  15848  103.4%  11
mem   15365  15808  102.9%  12
tor   14731  15071  102.3%  13
chi   14837  15100  101.8%  14
cle   14978  15207  101.5%  15

The Mavericks have had some injuries, but nothing that was really unforeseen. However, the trade of Josh Howard should have created a big shortfall in actual rotation minutes versus forecast. That's been offset by never-happened age declines predicted for Dirk Nowitzki, Shawn Marion and, especially, Jason Kidd, whose minutes are actually up over last season. Perhaps the whole idea of age-related, playing-time declines needs to be expunged from NBAPET. These numbers suggest that we more or less nailed Memphis' playing time projections but, in fact, that's not the case because NBAPET forecast more than 2,000 minutes for Allen Iverson. That's been made up for and then some by huge shortfalls in the playing time forecasts for O.J. Mayo, Zach Randolph, Marc Gasol, Rudy Gay and Michael Conley. The Grizzlies have just been a lot more stable than I anticipated, but that'll happen when you when.

No real surprises in Toronto. I thought Rasho Nestorovic would play more, but his over-projection was more than gobbled up by an under-projection of Andrea Bargnani that was based more on obsolete patterns than what he did last season. NBAPET did a solid job of anticipating down time for Luol Deng and Joakim Noah. However, Chicago's numbers don't include Taj Gibson, whom I did not forecast as a rotation player. He's going to end up with roughly three times as many minutes as NBAPET predicted. For Cleveland, Anthony Parker and Anderson Varejao have played more than I thought. Delonte West has played significantly less. Also, despite the month-long vacation Zydrunas Ilgauskas earned by being traded, waived and forced to sit, NBAPET is going to end up being very close on his projected minutes.

GROUP 4
TEAM  PROJ    ACT   %PROJ  RANK
mia   15154  15272  100.8%  16
orl   14872  14684   98.7%  17
sas   14167  13912   98.2%  18
lac   14837  14213   95.8%  19
uta   15084  14445   95.8%  20

GROUP 5
TEAM  PROJ    ACT   %PROJ  RANK
ind   13956  13131   94.1%  21
sac   15119  13858   91.7%  22
njn   14696  12631   86.0%  23
nyk   13956  11963   85.7%  24
cha   14273  11884   83.3%  25

The Pacers come up short because of T.J. Ford and Mike Dunleavy Jr., whose lack of playing time has not been accounted for in Brandon Rush's increased minutes. Also, the more-than-forecast minutes for Earl Watson don't fit in because he wasn't projected as a top-eight player for Indiana. The Kevin Martin injury and subsequent trade, as well as Francisco Garcia's physical woes account for the disparity in Sacramento. The rotations in New Jersey and New York have been a mess. The Raja Bell-for-Stephen Jackson swap accounts for most of the under-forecast for Charlotte, as does the disappointing season turned in by Tyson Chandler.

GROUP 6
TEAM  PROJ    ACT   %PROJ  RANK
gsw   13533  10910   80.6%  26
mil   14625  11789   80.6%  27
det   14590  11437   78.4%  28
was   15119  11387   75.3%  29
por   15683  11640   74.2%  30

Here, we have the bruised, battered and bewildering. The Warriors have become a way station for top D-League talent because of injuries that have resulted in an expanded roster for most of the season. My Bucks forecast was a disaster. My top three in projected minutes for Milwaukee: Joe Alexander, Hakim Warrick and Micheal Redd. If you want to step reading now, I understand. Injuries, age and ineffectiveness have all led to the erratic usage patterns in Detroit. The Wizards have seen their best player suspended and traded away the rest of their core. And then there's the Blazers. My preseason top eight included Greg Oden, Joel Przybilla, Travis Outlaw and Nicholas Batum. Despite huge under-forecasts for those injured (and traded, in the case of Outlaw) talents, the Blazers are on pace to win around 50 games. That's been one of the best stories of the season.

You can follow Bradford on Twitter at twitter.com/@bbdoolittle.

Bradford Doolittle is an author of Basketball Prospectus. You can contact Bradford by clicking here or click here to see Bradford's other articles.