Jeremy Lin says that the Knicks “may be the deepest team in the N.B.A.” (link). Is that true? I took the SSPM mid-season projections and screened for players who are projected to play at least 700 minutes and had a WARP48 of at least 0.067 (cutoffs chosen to maximize the number of Knicks). The Knicks had seven such players, which lands them near the top of the list. By this measure, Chicago is the deepest team in the NBA. Here are the teams with the deepest rosters.
CHI(9) (Derrick Rose, Luol Deng, Carlos Boozer, Ronnie Brewer, Joakim Noah, C.J. Watson, Kyle Korver, Taj Gibson, Richard Hamilton) SAS(8) (Tony Parker, Tim Duncan, Manu Ginobili, Kawhi Leonard, Matt Bonner, DeJuan Blair, Tiago Splitter, T.J. Ford) PHI(8) (Andre Iguodola, Louis Williams, Jrue Holiday, Thaddeus Young, Elton Brand, Spencer Hawes, Nikola Vucevic, Lavoy Allen) MIL(8) (Brandon Jennings, Andrew Bogut, Carlos Delfino, Beno Udrih, Ersan Ilyasova, Jon Leuer, Shaun Livingston, Drew Gooden) IND(8) (Danny Granger, David West, Roy Hibbert, Darren Collison, Paul George, George Hill, Tyler Hansbrough, A.J. Price) DAL(8) (Dirk Nowitzki, Jason Kidd, Jason Terry, Lamar Odom, Vince Carter, Shawn Marion, Delonte West, Rodrigue Beaubois) POR(7) (LaMarcus Aldridge, Gerald Wallace, Raymond Felton, Nicholas Batum, Jamal Crawford, Wesley Matthews, Marcus Camby) ORL(7) (Dwight Howard, Ryan Anderson, Jason Richardson, Hedo Turkoglu, J.J. Redick, Jameer Nelson, Glen Davis) NYK(7) (Carmelo Anthony, Amare Stoudemire, Landry Fields, J.R. Smith, Baron Davis, Tyson Chandler and Jeremy Lin). MIA(7) (LeBron James, Dwyane Wade, Chris Bosh, Shane Battier, Mario Chalmers, Mike Miller, James Jones) DEN(7) (Andrew Miller, Danilo Gallinari, Ty Lawson, Nene Hilario, Rudy Fernandez, Corey Brewer, Chris Andersen)
Zach Lowe’s excellent article explains that the max contracts that Dwight Howard andDeron Williams can sign are 6-yr, $138mm deals if they stay with their current teams, or 6-yr, $128mm if they leave. For the second number, he assumes that the players will resign with their new teams four years later at the then max rate, but the players would assume the risk of suffering a career-ending injury prior to resigning. At any rate, are Dwight Howard and Deron Williams worth $21-23mm a year? We project Howard (9 WARP per year) and Williams (8 WARP per year) will both be All-Star level performers for the next four years. Howard is probably underrated by our SSPM ratings, since our ratings favor players who control the ball. Win Shares are consistently higher than our SSPM WARP’s by around 3 wins per year, so even if we use their numbers, we project Howard will produce at most 12 WARP per year.
We argued that teams should maximize the number of players who deserve max contracts since the cap represents a possible market inefficiency. We applauded the Thunder when they locked up Russell Westbrook to a 5-yr, $80mm extension, so why are we hesitant to endorse the signings of Howard and Williams? First Westbrook is younger and improving (23 years old), while Howard (26 yrs old) and Williams (27 years old) are already in their primes. In contrast to Howard and Williams, we project Westbrook will average over 12 WARP over the next four years, and his average salary will be “only” $16mm. Howard’s minutes played, points and rebounds are all forecasted to trend down slightly over the next six years. Using $1mm + $1.6mm/WARP as our rule of thumb, Howard and Williams would need to generate 11 WARP per year to justify a $19mm year-one salary. Can they achieve 11 WARP per year? Possibly, but even in that case, the signing team would be getting at best fair market value, while assuming a lot of risk. The team would need other salary-efficient players to complete a championship-caliber squad.
Using the Statistical SPM ratings as explained in our NBA Chemistry paper.
MVP
LeBron James (8.2 WARP)
Rookie of the Year
Ricky Rubio (4.4 WARP)
Most Improved
Ryan Anderson
All-NBA
F: LeBron James F Kevin Durant C Dwight Howard G Chris Paul G Dwayne Wade
F: LaMarcus Aldridge F Pau Gaol C C Marc Gasol G Tony Parker G Kobe Bryant
Here are forecasted team WARP, assuming that rosters stay unchanged. The Sixers and Thunder have bright futures, while the Mavericks and Celtics have aging rosters.
Tm '12 '13 '14 '15 '16 1 MIA 57.0 51.9 46.6 41.7 36.7 2 CHI 54.7 51.4 48.2 44.4 40.4 3 OKC 45.7 46.8 47.8 48.3 48.6 4 PHI 44.5 47.6 49.7 50.6 50.2 5 DAL 42.7 36.1 29.7 23.2 17.2 6 ORL 41.9 38.1 34.0 29.7 25.6 7 POR 41.9 40.2 38.0 35.8 33.2 8 LAC 40.9 39.7 37.9 35.6 33.3 9 SAS 40.4 36.8 32.9 29.2 25.6 10 DEN 38.0 36.4 34.3 31.6 28.3 11 LAL 37.3 32.7 28.2 23.3 18.7 12 BOS 35.6 29.6 24.0 18.7 14.5 13 ATL 35.3 34.9 33.3 31.4 28.3 14 NYK 34.1 33.7 32.1 30.1 27.5 15 MIL 32.9 33.5 32.8 31.3 29.1 16 IND 32.6 31.8 30.2 28.2 25.4 17 MIN 32.4 34.7 36.5 37.9 38.2 18 MEM 32.1 32.3 31.4 29.7 27.5 19 HOU 31.4 32.4 32.7 31.6 30.0 20 UTA 23.8 25.5 26.4 26.2 25.5 21 GSW 23.0 25.9 28.0 28.5 28.0 22 PHO 21.0 21.4 21.0 19.9 18.0 23 NJN 19.7 23.5 26.2 26.9 26.9 24 NOH 19.5 20.9 21.1 20.5 19.4 25 SAC 17.7 22.2 25.8 28.3 29.8 26 DET 16.1 19.5 21.7 22.8 23.1 27 CLE 13.8 15.9 17.2 18.1 18.1 28 TOR 13.2 15.9 17.7 17.9 17.4 29 WAS 12.2 17.0 20.3 22.4 23.3 30 CHA 8.4 11.5 13.9 14.9 15.5
Our NBA Chemistry paper introduces the Skills Plus Minus framework which can generate mutually beneficial trades. One such trade would be a blockbuster exchange of two 26-yr old first time All-Stars: Portland’s LaMarcus Aldridge (6’11″, 22-8-3) and Chicago’s Luol Deng (6’8″ 16-7-3). Why does the framework think that this trade would benefit both teams? The simple answer is that Chicago has a surplus of defense, while Portland has a surplus of offense.
Chicago needs more inside scoring to complement Derrick Rose, and is already the best defensive team in the league. They can afford to part with Deng, who is the league’s best defender in terms of defensive RAPM. Portland, on the other hand, needs more defense, and has plenty of offensive-minded players (Crawford, Batum, andFelton) to replace Aldridge’s production. Would a frontcourt of Aldridge–Noah–Boozer be too big? Possibly, but the Bulls could arrange their substitution patterns such that two out of those three would always be on the court at any given time, withKorver or Brewer sliding into the small forward slot.
The salaries of the players match: Deng has 3-yrs, $40mm left on his contract, while Aldridge has 4-yrs, $54mm.
Long before Jeremy Lin and Linsanity, basketball has been a favorite pastime of the Asian American community dating as far back as the early 1980s. While Asian Americans may not be the most desirable ethnicity of a potential basketball teammate on the playground (see: http://www.youtube.com/watch?v=qiektV3HEB8), people should not be so quick to dismiss the talent in these players. The rising popularity and availability of youth basketball programs in Asian American communities have helped propel many recreational and streetball players to become successful high school and collegiate athletes. Your typical all-Asian American basketball team consists of four players under 6 feet and one 6+ footer who will play the “center.” Though not intimidating at first glance, they will most likely be scrappy on defense, push the ball whenever they get the chance, draw fouls on aggressive drives to the basket, and shoot the lights out from downtown. Every so often, you will have an Asian kid who throws down a dunk on a breakaway layup.
The New York Times wrote an article about my old basketball team, the New York Rockits, and a recent Asian basketball tournament that they held over the weekend.
For more information on Asian basketball leagues and tournaments in the NYC area, see the below websites:
www.nacbait.com
www.fastbreaknyc.com
www.asianhoops.com
www.dreamleague.org
Best Performances
Derrick Rose (0.340 WARP, 32-2-9, 3 stl, 2 blk, 2tov). I bet he wishes he had some offensive support from, say, LaMarcus Aldridge.
Paul George (0.329 WARP, 11-9-5, 4 stl, 1 blk, 1 tov).
Worst Performances
Jerryd Bayless (-0.207 WARP, 4-2-2, 5 tov in 11 MP)
Monta Ellis (-0.192 WARP, 14-5-2, 5 tov, 7-19 FG, 0 FTA). Rumor that Orlando is trading Superman for Ellis???
Vince Carter (-0.167 WARP, 2-1-0, 2 tov in 22 MP)
Penn upset Harvard 55-54 last night and pull within a half-game of the Crimson. Kyle Casey scored a go-ahead layup with 5.7 seconds left, but the basket was nullified by a controversial charging call. The cheating is worse than ever!! At least the hockey team won, finishing #3 in the ECAC and earning a bye in the ECAC playoffs.
Best Performances
Brandyn Curry (0.235 WARP, 5-1-5, 2 stl, 1 blk)
Wesley Saunders (0. 182 WARP, 10-2-2, 5-6FG). Saunders, Harvard’s 4-star freshman, had a nice game off the bench. One of Harvard’s most heralded recruits, he has been overshadowed by fellow freshman Steve Moundou-Missi most of the year.
With the trade deadline approaching, now is a good time to ask ourselves how much money should teams pay per win? According to Hoophype, the average NBA team payroll is $63.6mm. Analysis from our NBA Chemistry paper shows that a team of “replacement level” players will win around ten games per season. That means that the average NBA team will win 41-10 = 31 games more than a team with all replacement players. The minimum salary in the NBA is roughly $1mm per player (depending on years in the league). Since there are 15 roster slots per team, that means that the average available payroll per team is roughly $63.6mm-15*$1mm = $48.6mm. Therefore, on average, teams should pay on average $48.6mm/31 = $1.6mm per WARP (“win above replacement player”).
There are (at least) two inefficiencies in the NBA labor market:
* Players whose contracts are dictated by the rookie scale. Blake Griffin, for example, is making $5.7mm this year, even though he is expected to generate 9-10 wins. That’s bargain!
* Players making the maximum contract. LeBron James is making “only” $16.0mm, even though he has been averaging nearly 20 wins per season.
If you take away these two pools of underpaid players, there are more dollars left over for the remaining players. So in equilibrium, teams will pay more than $1.6mm per win for the remaining players. Do they? Using the SPM player projections, I regressed salary versus expected wins for players no longer under the rookie scale, and found a slope coefficient of $1.4mm.
To optimize payroll efficiency, general managers should therefore maximize the number of players under the rookie scale and players who deserve maximum contracts. For other players, teams should limit the annual salaries to the minimum plus $1.4mm-$1.6mm per expected win. If they follow these guidelines, they will sure to build a winning roster.