ESPN NBA

Before Allan Maymin, Philip Maymin, and Eugene Shen introduce their new basketball metric, "Skills Plus Minus," they conjure up the spirit of Steve Jobs, quoting the Apple founder at the top of their paper:

My model for business is The Beatles. They were four guys who kept each other’s negative tendencies in check. And the total was greater than the sum of the parts. Great things in business are not done by one person; they are done by a team of people.

To borrow from the parlance of hoops, John Lennon, Paul McCartney, George Harrison and Ringo Starr might not have been the best musicians at their respective positions, but their individual skill sets created magic because those skills complemented each other perfectly.

Harrison's restraint as a guitarist was a perfect shadow for McCartney's sunniness. And Starr's devotion to tempo (even weird time-signatures like "All You Need is Love") punctuated Lennon's moony vocals.

That's called good chemistry and it's the prism through which the authors of "NBA Chemistry: Positive and Negative Synergies in Basketball" look at how NBA teams can make beautiful music together on the court.

Measuring chemistry is tricky
Folks who use hard data to measure success and failure in sports are skittish about ascribing losses to bad chemistry or chalking up winning to good chemistry. That's because chemistry always seems to be a chicken-and-egg game marred by tautology:

Good teams have chemistry. How do we know? Because they're winning -- and winning is the product of good chemistry.

Yet even the most hard-core analysts have an inkling that there are certain players and skills that optimize each other on the floor. You're smart to surround Dwight Howard with perimeter shooters, and smart to pair Chris Paul with a big man who can punish defenses in the pick-and-roll.

Those aren't advanced discoveries -- just intelligent observations from watching the Magic lead the world in 3-pointers made over the past few seasons and from seeing David West drain face-up jumper after face-up jumper as defenses try to trap Paul.

An eye test is one thing, but hard data is another. What if we could identify less obvious skills (and the players who embody those specific skills) that, if placed alongside each other on the court, could improve your team's chance of winning basketball games? Is loading up a lineup with the five best available players always the best idea? Could piecing together a unit with specialists create better synergy, even if those specialists might be lesser overall players in our minds?

George Harrison for Jimmy Page?
Would the Beatles have been better off swapping out George Harrison for Led Zeppelin's Jimmy Page -- considered by many music junkies to be one of the best rock guitarists ever -- or was Harrison's less dynamic, more understated style a better fit for what the Beatles wanted to accomplish?

We'll never know, but it's hard to find fault with the Beatles' oeuvre. In that same spirit, you could argue that Led Zeppelin would've been worse off with Harrison instead of Page.

In other words, Harrison-for-Page would've been "a trade that hurt both bands."

But how is that possible? If one guitarist -- or point guard, or center, or small forward -- is empirically better than the other, then wouldn't he be a more valuable member of the band or team, no matter what the circumstances?

The essence of Skills Plus Minus -- and the CP3/D-Will debate
In search of some answers, the authors, all of whom work as quantitative traders in the financial world, put their new metric to the test. They describe the basic parameters of Skills Plus Minus:

We introduce a novel Skills Plus Minus (“SPM”) framework to measure on-court chemistry in basketball. First, we evaluate each player’s offense and defense in the SPM framework based on three basic categories of skills: scoring, rebounding, and ball-handling. We then simulate games using the skill ratings of the ten players on the court. The results of the simulations measure the effectiveness of individual players as well as the 5-player lineup, so we can then calculate the synergies of each NBA team by comparing their 5-player lineup’s effectiveness to the “sum-of-the-parts.” We find that these synergies can be large and meaningful. Because skills have different synergies with other skills, our framework predicts that a player’s value is dependent on the other nine players on the court. Therefore, the desirability of a free agent depends on the players currently on the roster. Indeed, our framework is able to generate mutually beneficial trades between teams.

As Shen says, "A player's value to his team depends on the skill of the other players on that team." A sophisticated metric like Advanced Plus Minus (and Regularized Advanced Plus Minus) have been inordinately useful, particularly in identifying which 5-man units work well together. Skills Plus Minus builds on that work by trying to answer the question, "Why?"

What is it about Player X's skills that make him a better fit in a lineup when you sub him in for Player Y, who happens to be regarded as a better overall player?

Unlike Advanced Plus Minus, Skills Plus Minus simulates possessions to account for variables (a possession that starts with a steal produces different results from, say, a possession that begins out-of-bounds). Looking at player attributes is an important ingredient in answering many of these questions. (For an advanced illustration, please visit Dean Oliver's study on beach paddleball, a sport which not only serves as a nice alternative to body-surfing or the construction of sand castles, but offers a good laboratory for this kind of examination.)

In SPM's first case study in their study, the authors tackle one of the more spirited debates in recent years: Would the Hornets or Jazz have won a Chris Paul-for-Deron Williams trade in the summer of 2010? Their discovery:

[S]urprisingly, the answer is YES to both. A Williams-for-Paul swap would have made both teams better off and is an example of a mutually beneficial trade.

The authors broke down Paul and Williams' games using those three basic categories on both ends of the floor: scoring, rebounding and ball-handling. After the data was sufficiently crunched, the findings were interesting:

A Paul-for-Williams swap would've helped both teams.

"We thought that was pretty interesting," Shen says. "It turns out our framework predicts that stealing the ball has positive synergy. So if you have two guys on your team who steal the ball, it will actually generate more steals than if the two players played separately. Chris Paul steals the ball very well -- but his teammates do not. But Utah does. On the flip side, Deron Williams would've fit better on New Orleans because our framework predicts that offensive scoring has a negative synergies because you have to share the ball. On the Hornets, Williams wouldn't have to share the ball with as many teammates."

Which skills are good fits?
The research team pored over a ton of data, ran countless simulations and looked at how many points certain combinations of skills created.

Some of the conclusions are obvious to the basketball fan: Offensive ball-handling and offensive scoring have positive synergies. Likewise, offensive ball-handling has a positive synergy with offensive rebounding because, as the paper explains, "[O]ffensive ballhandling helps a team convert possessions into shot attempts, and offensive rebounding increases the number of possessions over which the ballhandler can protect the ball." In other words, give Chris Paul teammates who can score and others who can generate more possessions on the offensive glass, and his team will thrive.

But offensive ball-handling has a negative synergy with itself. It's not that having a couple of guys who can capablly handle the ball is a bad thing, but your team simply won't be able to extract the full value of those skills because there's only one basketball on the court. If Ball-handling stud A has possession of the ball, by definition Ball-handling stud B doesn't -- and that minimizes his best skill.

Here's a result that initially surprised me: Offensive rebounding has a negative synergy with offensive scoring. That seems counterintuitive, until Shen explained it to me.

"Players who have trouble scoring generate more missed shots than players with high offensive scoring ratings," Shen says. "So offensive rebounding will be more valuable to a team that misses shots because there are more opportunities."

One pattern that emerged was that "rare events" (like steals/defensive ball-handling) tended to produce positive synergies, while "common events" (like defensive rebounds) produce negative synergies. How come? Because increasing a team's rebounding rate from 70 percent of defensive rebounds (which would be lousy) to, say, 75 percent (very good) represents only a 7 percent increase. But upping offensive rebounds, which aren't nearly as common as defensive rebounds, from a rate of 30 percent to 35 percent represents a robust 17 percent gain.

The mutually beneficial trades
In addition to the Paul-for-Williams deal, the authors of the study identified 222 potential trades among starters in 2010 that could've helped both teams. Although that seems like a lot, 222 represents only 2 percent of all possible trades.

Some of these trade scenarios, like Paul-for-Williams, are fascinating to contemplate. Others, like Marvin Williams for Daequan Cook, elicit nothing more than a yawn. But on a few rare occasions, the proposed trade is mind-boggling, like sending Amare Stoudemire from Phoenix to Minnesota for Ryan Gomes a couple of years back -- a trade the system says would've helped both the Timberwolves and the Suns at the time.

I asked Shen whether, as a scientist, he was genuinely happy to see the system produce such a novel result or was it frustrating because it might undermine the credibility of the study.

"Amare Stoudemire's reputation is so much better than Ryan Gomes', so the first reaction is 'Can that be right?' Shen says. "But I'd be dishonest with myself if I rigged the system so it doesn't spit that out. For what it's worth, Amare Stoudemire's ratings are not very high. He's rated very close to Ryan Gomes [between 2006-2010]. They have very different skill sets, and the system predicts they would've been a good trading fit. So, I guess, I find it interesting and thought-provoking that the system would generate a trade like that. But on the surface, of course, everyone thinks Stoudemire is better and that would never happen in real life. But I hope it stimulates some good discussion."

I told Gomes about the paper, which not only had him in a mutually beneficial trade for Stoudemire, but also Luis Scola and Udonis Haslem. Gomes got a good laugh out of it but, like Shen, maintained that specific attributes don't get emphasized enough in the NBA.

"Different guys have different skills," Gomes says. "You see it all the time. You might be a bad fit on one team, get dealt, then all of the sudden you play great under a new system with new teammates and a new coach."

The takeaway
Several smart people I communicated with who work with advanced stats had a generally favorable impression of the paper -- though none of them would deal Stoudemire for Gomes. But they unanimously praised the effort and feel that the study represents a natural progression in the discussion of analytics.

Figuring out the component parts of what we know as chemistry or synergy is one of the next great frontiers of this movement. It's not enough to put an exceptional distributor on the floor. To maximize that point guard's gifts, a team must surround him with the right combination of players -- and that combination might not always be the sexiest free agents on the market.

For all his offensive failings, rebounding machine Reggie Evans might be the perfect power forward for a dime-and-steal happy point guard. And is it possible that Thabo Sefolosha, who ranks second among starting 2-guards in defensive rebounding, helps offensive scoring juggernauts Kevin Durant and Russell Westbrook in ways that aren't apparent to the naked eye?

In an NBA where the margins of victory are razor thin, every variable counts. And it's becoming increasingly clear that things we've always regarded as assets or liabilities reside in a very gray area.

After taking three overtimes to decide Game 4, Wednesday’s Game 5 between the Oklahoma City Thunder and Memphis Grizzlies was pretty much over after three quarters. The Thunder entered the fourth quarter with a 19-point lead and ended up taking a 3-2 series lead with a decisive 27-point victory. It's the largest margin of victory for either team this postseason.

The Thunder became the first team in four tries since 1993 to win their next playoff game after winning a triple OT postseason contest. Another victory would make Oklahoma City the first team since the 1976 Boston Celtics to win a series after winning a triple-overtime playoff game.

The Thunder held Memphis to 20-or-fewer points in every quarter. The Elias Sports Bureau tells us that was the first time in eight years that the Oklahoma City franchise did not permit more than 20 points in any quarter of a regular-season or playoff game. That had last happened on March 24, 2003, when the Sonics beat the Phoenix Suns, 82-73.

Two keys to the Thunder’s victory were transition points and bench production.
The Thunder entered Game 5 having outscored the Grizzlies 70-44 in transition during the first four games of the series, including 40-16 in games 2 and 4 (both Thunder wins), and it was more of the same for Oklahoma City on Wednesday. The Thunder shot 78.6 percent and didn’t turn the ball over once in transition.

The Thunder bench outscored the Grizzlies bench, 53-27.

A huge part of that production was James Harden. After posting a combined plus-minus of -8 in games 1-4 of the series, Harden had a game-high plus-minus of +29 in Game 5. Harden provided a notable boost off the bench as the Thunder shot 51.9 percent (27-52) from the field while Harden was on the floor, compared to just 38.4 percent (10-26) when he was not. Harden may be in line to receive more minutes in Game 6 as the man he replaced in the lineup, Thabo Sefolosha, had a team-low plus/minus of -2 in Game 5, the only member of the Thunder to net a plus/minus under zero.

The Thunder move to within a win of their first Conference Finals trip since 1996, when they were the SuperSonics and they reached the NBA Finals, losing to Michael Jordan and the Chicago Bulls.

What are their chances? In NBA playoff history, when teams are tied 2-2 in a best-of-seven series, the Game 5 winner goes on to win the series 83 percent of the time.

The 2010 baseball season was highlighted by a pair of perfect games.

In this NBA season, perfect efforts are coming in bunches, if we're reasonably loose in our definition of the term.

Thabo Sefolosha was 4-for-4 from the field and 3-for-3 from the free throw line, good for 13 points as the Oklahoma City Thunder came back to beat the Hornets on Monday night.

That caps a 10-day span in which we've seen some impressive perfection performances. On November 19, Los Angeles Lakers reserve Matt Barnes was 7-for-7 from the field, 5-for-5 from 3-point range and 5-for-5 at the foul line in a win over the Minnesota Timberwolves.

The next day, DeShawn Stevenson of the Dallas Mavericks was 4-for-4 from the field, including 3-for-3 from three, and two-of two at the foul line in a victory over the Atlanta Hawks.

Then, on the 21st, Lakers forward Pau Gasol went 10-for-10 from the field and 8-for-8 at the line in a win over the Golden State Warriors.

A check of the Basketball-Reference.com Play Index indicates that Sefolosha was the 11th player this season to make at least four field goals in a game AND be perfect from the free throw line. But he, Barnes, and Stevenson are the only three in the NBA whose perfection effort included multiple 3-point field goals.

Over the last couple of seasons, this sort of thing is a five-to-six times a year occurence, but in the small sample size that is the early part of this schedule, we're on pace for significantly more than that.

Also of note from a perfection perspective in this game was the performance of Thunder center Serge Ibaka, who was coming off a perfect 7-for-7 in a loss to the Rockets on Sunday. Ibaka was perfect from the field, but imperfect otherwise in a one-point loss. When he was on the floor, the Thunder were outscored by five points.

Monday, Ibaka learned that imperfection was acceptable On a night where Russell Westbrook and Kevin Durant starred, Ibaka was 0-for-5 from the field in 30 minutes off the bench, BUT when he was on the court, the Thunder outscored the Hornets by 15 points. No Thunder player had a better plus-minus than that.

Prior to Saturday night's game between Cleveland and Oklahoma City, the Cleveland Plain-Dealer's Brian Windhorst wrote:

I believe the Durant-LeBron matchup could be better than the Kobe-LeBron matchup tonight and for years to come. Durant is a high efficiency scorer, he gets to the line at a high rate, shoots at a high rate and he rebounds better than Kobe. At this point in his career, at least this season because of injuries, Kobe has turned into a volume scorer on a lot of nights. Durant, and James for that matter, aren't. That is why I predict a quality duel.

True to Windhorst's forecast, the mano-a-mano battle was captivating. The two scorers combined for 71 points in a seesaw affair that saw five lead changes in a scintillating fourth quarter. A battle that was waged in the interior for three quarters -- and dominated by Shaquille O'Neal for much of that time -- moved further out to the perimeter in the fourth.

For all of Durant's uptick in efficiency (a PER of 24.58, vs. 20.85 last season), the Thunder have been winning basketball games this season on the strength of their defense, and we saw Oklahoma City make some gritty stands down the stretch. There's a reason it took some huge shot-making by Daniel Gibson for the Cavs to put Oklahoma City away at the Q -- the Thunder clogged the middle, as they've been doing all season.

But just as we're not hearing enough about Oklahoma City's defense amidst the celebratory praise of everyone's favorite youth movement, there's not much discussion of this:

The Thunder have trouble scoring points.

In fact, only nine teams in the League have more trouble.

While that's a marked improvement from last season, when only the Clippers were worse than the Thunder in offensive efficiency, it's a little bit of a head-scratcher for a team blessed with a matchup nightmare like Durant.

Saturday night during the tight fourth quarter, we got a glimpse of the Thunder's struggles when they went more than five minutes without draining a shot from the field. There was a particularly ghastly stretch of seven possessions over which Oklahoma City generated only a single point on a Durant free throw.

What happened to the Thunder in those moments? Was Durant not finding shots he likes? Was it something akin to what the Lakers or Cavs experience at times when the other four guys on the floor stand around watching Bryant or James? Was Durant forcing the issue? Not forcing it enough?

Possession 1 (5:35)
Inefficient offensive units often have a tendency to squander a good 10 seconds before getting into their sets. Finding good shots against a defense as stingy as Cleveland's is a tough business, and the more time you budget to generate those looks the better. By milking 10 seconds off the clock, you also let your opponent off the hook because it requires far less energy to defend for 14 seconds than 24 seconds.

On this possession, rookie combo guard James Harden has the ball up top. It appears that the Thunder might be running a pin-down with Russell Westbrook on the right side to free up Durant, but if that's the case, Westbrook misses Durant's defender, Anthony Parker, altogether. When Durant gets the pass up top and begins working against Parker, James leaves Harden to double team. No surprise there.

Durant, sometimes criticized for being an unwilling passer, kicks the ball out to Harden, who passes up the 3-pointer (he's a 37.7 percent shooter from that distance). Harden instead works off the dribble, but it isn't long before he tosses the grenade back to Durant with the clock expiring. Durant has to settle for a long, contested shot from beyond the arc:



Possession 2 (4:46)
Another half-hearted down screen for Durant, this time by Jeff Green. Now might be a good time to send some film to the Thunder supporting cast of Kendrick Perkins laying out for Boston's perimeter scorers. Space matters, and the more room a team can generate for its primary scorer to work, the more efficient that offense is going to run. Fortunately for Durant, Serge Ibaka gets himself between Durant and Parker. This gives Durant one of his better looks at the basket in the fourth quarter, though it's not wide open. Why not?

Check out Shaquille O'Neal! You won't see him step up to challenge a shooter on a pick-and-roll very often, but here he sticks a big limb in Durant's face:



Possession 3 (4:23)
It's not a coincidence that Oklahoma City draws a foul early in the possession. Notice how much more quickly and decisively they challenge Cleveland, as Durant makes Parker chase him from the moment they cross the time line?

After the inbound, though, the Thunder have a difficult time freeing up Durant. Throw some credit Anthony Parker's way. Time and time again in the fourth quarter, he dodges Thunder picks, not yielding an inch to Durant. There's a telling moment at the 10-second mark. Watch:



See how Westbrook picks up his dribble? He assumes that he'll lob a simple entry pass to Durant, but Parker is doing such a good job denying that pass that Westbrook has to swing the ball over to Sefolosha in order to get his dribble back on the return. At that point, Westbrook has to freelance, and Daniel Gibson -- yes that Daniel Gibson -- blocks his runner as the clock expires.

(Read full post)

One of the simple ways of experiencing basketball is by talking about it with people who share your love of the game. One of the people I enjoy rapping with is John Krolik of Cavs the Blog and SLAM Online. The best conversations are the ones that produce interesting questions, then aim to answer them. Here are some of those questions about the NBA John and I have been bouncing around in our last couple of conversations:

As guys get freakier and more athletic, are we witnessing an end to positional orthodoxy?
JK: We're definitely seeing a lot of blurring in positional lines, particularly outside of the center position. One thing in particular I like is the rise of the true combo guard. Early in the decade, we got a lot of alleged "combo guards" who were really just superpowered bench gunners given control of teams with mixed results; Stephon Marbury, Steve Francis, et cetera. (Iverson is Iverson.)

But now we're really starting to see effective players who are a cross between the one and the two in a good way, and they're being complimented with other multi-skilled guards rather than going with a strict point guard/shooting guard backcourt. In San Antonio, they put Tony Parker, who's a great scorer for a point, next to Manu, who's a great playmaker for a shooting guard, and things went well. The double-combo backcourt of Mo Williams and Delonte West turned Cleveland's backcourt from a disaster area to a huge strength last season. Even Jason Kidd, the truest of points, is playing with JET and JJ Barea, and has even become adept at knocking down catch-and-shoot 3s off of other people's assists. Phil Jackson's won only 10 championships using an offense that doesn't require a traditional point. And so many young combo guards are coming in with tons of talent: Tyreke Evans, Russell Westbrook, Brandon Jennings and even John Wall, who should definitely be put next to a guy who can pass and shoot when he comes into the league so that he can spend some time in each game going on guilt-free scoring rampages. Wall might be the combo-guard messiah.

KA: This is a beautiful trend because it's created a much more diverse range of basketball styles. Very few teams around the league look alike, even though many of them run much of the same stuff. The fact that so many players can do so many different things on the floor creates an exponentially greater number of things a team can do schematically. On many teams, shots on the floor can be drawn up for almost any player at any spot! Part of this can be attributed to athleticism. One the things that made a power forward or a center a big men was his ability to perform big men tasks -- rebounding, shot-blocking, the ability to routinely get high-percentage shots close to the rim. Today's NBA perimeter players have the athleticism to do a lot of that -- and many of the bigger guys in the league have perimeter skills, as well.

This seems like a nice segue to ...

Do traditional big men have a future?
KA: Whether you chalk it up to the prohibition of hand-checking or the stylings of Mike D'Antoni's Phoenix Suns teams (I'd argue that former rendered the latter), the professional game has undergone a seismic shift over the past decade. Perimeter play has taken over. Today's power forwards have big guard games and two of the top three players in 3-point attempts are 6-foot-10. It's a world gone mad, but you can't complain about the product on the court. The NBA has never been more fun to watch, and we're just getting started...

...or are we?

Trends have a way of feeling permanent while they're being experienced, but they rarely last forever. At some point, laws of macroeconomics take over. Right now, there aren't more than a handful of big men in basketball who have refined post moves and can drain a running right-handed hook with consistency. Teams don't value those attributes as much as speed and 3-point shooting. But as more and more players have the ability to drain 100-200 3-pointers per season at a 40 percent clip, the demand will shift. Kids who arrive on the NBA's doorstep with the ability to dominate the game inside with uncanny efficiency will be shopping skills that few teams will be able to defend.

JK: I'd say the hand-check rules imposed an artificial set of circumstances that forced a change, so I don't think we'll see the pendulum swing all the way back to where it was. But I think guys are finding out that even though big men need to be faster and more skilled than they used to be and can't count on getting minutes just because they can score with their backs to the basket and do nothing else (i.e. Eddy Curry), the post-up game is still a valuable weapon. Look at the Lakers. Andrew Bynum, when he's engaged, defends the rim, gets rebounds and is quick enough to find room and finish off of others, but also posts up. Pau Gasol plays the high-post, runs the floor, gets rebounds, passes beautifully and can knock down the mid-range jumper, but also has a wonderful post game. And of course Kobe can and does do just about anything that's possible for a basketball player to do, but also utilizes the post game.

I'd say that the post-up specialist won't be in vogue again in the foreseeable future, but more and more bigs and wings who can do what's demanded of them in the post hand-check NBA are going to find that the actual post game is still a hugely valuable weapon, especially as fewer and fewer teams know how to defend it.

Of the current young up-and-coming teams, which ones are for real and which ones will provide an entertaining illusion of success?
KA: When sizing up a team's future prospects, the first thing I ask myself is, "Can I imagine this team ranking in the top half of the league defensively?"

Oklahoma City is the quintessential upstart squad. They're fun, charismatic, dynamic, athletic ... and not all that impressive as an offensive unit. It's the Thunder's defense that's led them to a 17-14 record this season. So long as tough, lanky defenders like Russell Westbrook and Thabo Sefolosha are patrolling the perimeter (and James Harden too), opponents are going to have a tough time scoring against them. With that Kevin Durant angle pick-and-roll as the anchor of their offense, they're a good bet to win a playoff series sometime soon.

Brandon Jennings has sparked any and all attention the Bucks have received this season, but Milwaukee's frontcourt of Andrew Bogut, Ersan Ilyasova and Luc Mbah a Moute have put up gritty defensive numbers. Mbah a Moute comes as no surprise, but I was shocked by Bogut's stats, until I looked at his figures under Scott Skiles last season -- also really, really good. Once they get a (healthy) shooting guard who can play drive-and-kick off the Jennings-Bogut pick-and-roll, the Bucks could be dangerous under a coach who was booted from his last gig in Chicago after assembling the league's top-ranked defense and the Eastern Conference's 3rd best record the previous season.

Sacramento's lousy defensive numbers don't concern me right now. They strike me as a team that's going to experience a major overhaul over the next 18 months, and a big part of that metamorphosis will be acquiring some pieces around Tyreke Evans who can defend. I have less faith in Memphis, Minnesota, Golden State and, to a slightly lesser extent, Philadelphia, who all have rosters riddled with defensive ciphers.

JK: I think Oklahoma City wins a playoff series when their backcourt clicks into place, and that's close to happening. I love Westbrook's game and think he has a ton of potential, but he just needs to be more disciplined. He pushes the ball, plays great defense, and does all these little things, but then he'll throw up a bad jumper, brick a full-speed reverse layup, or make a silly pass, and his true shooting percentage and turnover rates are way off of where they need to be because of that. It'll be interesting to see if the answer there is Harden maturing to the point where he can play 30-35 minutes a game and cover some of Westbrook's weaknesses with his shooting, playmaking and ability to create off the drive. (Combo guards!) But I think that young frontcourt is the envy of a lot of teams in the league, Sam Presti keeps getting valuable pieces without giving up much, and I'd call the future very bright there.

For Sacramento, the short-term question is how Tyreke is going to work with Kevin Martin. They might cancel each other out or become absolutely unstoppable together, although they might need to do the latter to make up for Martin's suspect defense. But Thompson, Hawes, Casspi, and even Brockman all look like keepers, and Tyreke has given every indication that he can be built around.

In Milwaukee, I think they should be having serious brainstorms on how they can hide Mbah a Moute on offense so they can keep him on the floor longer, maybe even looking for a stretch four so they can put Mbah a Moute closer to the basket offensively and use him like Detroit used Ben Wallace. He's that good defensively.

I agree with you about the rest of the teams, although I give Memphis some upside because I think it's a bit too early to completely give up on Hasheem Thabeet as an impact player defensively; if Orlando could build a defense around Howard and four perimeter guys, there's a chance Memphis can as well. (A chance, mind you.)

What is it about Stan Van Gundy that we like so much?
JK: I think we've got a pretty narrow view of how to evaluate coaches, because we don't see the vast majority of what they do and we're trained to look for their failures and not their successes. Coaches almost exist to be fired, and every time they make a mistake with their play-call or substitution, it'll get talked about the next day.

I think the biggest job of a coach isn't to call timeouts strategically or be a genius with his in game substitutions. (Although both are definitely important, especially the latter.) I think the job of an NBA coach is to set up a system that best utilizes the talents he has available to him, and that's where Stan Van Gundy comes in, especially last season. Of his five starters, he had three guys with below-average defensive reputations, Dwight Howard, and a rookie.

Instead of trying to have everyone play straight-up or stick Rashard Lewis at the three, he evaluated what he had -- the best shot-blocker in the league and more quickness on the perimeter than most other teams had. So he stuck Lewis at the 4 and never looked back, and built a defense around running other teams off threes and keeping Howard at home under the basket. What happened? The Magic gave up the second fewest made baskets at the rim, the second fewest made 3s per game, and more shots from 10-15 feet and 16-23 feet than any other team in the league. They also had one of the league's three best defenses in terms of efficiency.

Offensively, he had Dwight Howard, who can catch and finish with the best of them but isn't a great post player, more shooting and playmaking at the forward spots than most anyone, and a bunch of guys who can shoot threes. So he had Howard look for catches at the rim, ran 3/4 screen-rolls, and had his players shoot a bunch of threes rather than try to do what everyone else was doing. Van Gundy's failures last season were there for the world to see, but what he did extremely well was more subtle.

KA: I like his press conferences, too. The irony of Van Gundy is that popular perception sometimes paints him as inflexible. But as you said, no coach sculpted a more sensible system for his personnel last season than Van Gundy. He did a full appraisal of his talent, saw where he had edges over his opponents at each position (ballhanding at the 3, shooting at the 4, mobility at the 5) and designed his offense to exploit those advantages.

This isn't to say there's anything wrong with building an elite team by first implementing the system, then by populating that system with players whose talents most conform to it. Whatever works, by all means. Just win. But the ability to create a system around a disparate collection of talent that was brought together randomly is in many ways even more impressive.

Should LeBron James be playing more power forward?
KA: Despite James’ size, strength and efficiency on the glass, Mike Brown has him firmly situated at the small forward slot. In fact, you have to go pretty far down the list of Cleveland’s 5-man lineups to find units in which James is playing power forward. But in the six lineups that feature James surrounded by one traditional big man and three smaller players for at least 10 minutes, the Cavs outscore their opponents 96-83 (prorated for 48 minutes).

Those numbers are enough for me, but let’s think about it in practical terms. We’ve already discussed how positional dogma is a thing of the past in an NBA that’s much smaller than it was 10 years ago. When thinking about how to best maximize LeBron in the half-court, wouldn't you prefer that he drag a bigger defender out to him in order to create more space on the floor for your offense? And defensively, wouldn’t a team like Cleveland, whose primary weakness has been its plodding frontcourt, be better served by having LeBron cover Rashard Lewis on Orlando’s pick-and-pop or Boston’s bigs on the Celtics’ rotating screen-and-rolls? Doesn’t it make more sense to challenge Stan Van Gundy and Doc Rivers to match up with a more athletic lineup? And wouldn’t Cleveland benefit from more transition opportunities?

Would team rebounding suffer? When you look at those aforementioned six lineups with LeBron at the 4, the answer is no. Apart from the political stickiness of limiting the minutes of the Cavs' veteran big men, I have trouble seeing how making the Cavs a more athletic team around LeBron comes with much downside.

JK: The short answer is that I'm extremely confused as to why LeBron doesn't get more time at the 4 position, at least for around 10 minutes of his time on the floor. I understand some of the reasoning behind not giving him significant minutes down there. The Cavs show hard on every perimeter screen, which would require LeBron expending more energy on the defensive end than the Cavs are comfortable with, especially in the first three quarters. And of course, the Cavs don't want LeBron in foul trouble under any circumstances. And generally speaking, the Cavs' big men are better players than Jamario Moon, who typically plays the 3 in the Cavs' small-ball lineup. But LeBron getting the ball in the 10-15 foot range and making his move from down there is absolutely deadly, and that small-ball lineup should definitely be something used more often to keep opposing teams on their toes.

What confuses me more than anything is that while the Shaq/Varejao frontcourt has some offensive issues and the Shaq/Hickson frontcourt has some serious defensive issues, a Shaq/LeBron frontcourt hasn't been tried at all this season, and I mean at all. I suppose the reasoning is that LeBron would be forced to expend way too much energy on the perimeter defensively as Shaq sags to the paint on pick-and-rolls (LeBron's never gotten minutes at the four alongside Z either), but with the Cavs supposedly looking for a "stretch 4" at the deadline to make life easier for Shaq, it's odd that they haven't at least tried using LeBron in that role.

How Real is Dallas?
KA: Little known fact: Of the 50 5-man units that have played together the most this season, two of the top three in overall efficiency belong to the Dallas Mavericks. Whether it's Jason Terry or J.J. Barea at the shooting guard, the Mavs' big names are absolutely crushing their opponents on both ends of the floor. Dallas is a Top 5 defensive squad and features one of the game's great shotmakers in Dirk Nowitzki. They also have tremendous flexibility to match up with opponents on either end. They can play old-school or new-school. Want to tease the Mavs with small ball? That's fine, because they're perfectly good going with three guards and moving Shawn Marion and Nowitzki into the frontcourt. Want to try to outmuscle them? Erick Dampier may have an outsized contract, but he's also one of the better basket protectors and garbage collectors in the league. Opponents shoot a measly 57.4 percent at the rim against the Mavs -- only Boston, Cleveland and San Antonio are better.

More than anything, the Mavs strike me as a team composed of professionals. These are serious basketball players led by a serious coach. Is it possible that a squad with so many thirtysomethings breaks down physically over the course of an 82-game season? Perhaps. But where some see brittleness, I see experience. In fact, I see shades of the best San Antonio Spurs squads. I see a team that truly understands its collective talents and limitations and puts a premium on execution.

Can they compete with the Lakers in late May? I'm not sure anyone in the Western Conference can, but Dallas -- with its length, smarts, and perimeter prowess -- might just be the toughest competition the Lakers encounter.

JK: Dallas has a ton of talent, Dirk is right up there with the best players in the league, and the team defends. My caveat would be that they're thinner than people think, and much more dependent on Dirk. As of December 26th, Dallas was +11.6 points per 100 possessions with Dirk on the floor and a stunning -16.5 points per 100 with Dirk on the bench. As bad as LeBron and Kobe's benches are, their teams are only -8 when they sit, to offer some perspective.

A lot of that has to do with Drew Gooden; Gooden's plus-minus is -23.1, and as someone who's watched a good deal of Gooden in his life, I can tell you that's not random noise. Drew Gooden is the anti-Battier. I'm also not a huge J.J. Barea fan. He's fun to watch and works fairly well with Kidd offensively, but I believe you were the one who said he plays defense "like a man frantically searching for his car keys," and the plus-minus numbers support the theory that Barea's somewhat of a defensive liability. Dallas can play with anyone, especially when Dirk's on the floor, and if they do something to get a better backup for Dirk than Gooden and hide Barea's defense a little better (maybe play more Beaubois, who's gone through growing pains and will probably continue to do so, but has lockdown defensive potential), I'd call them a true force to be reckoned with in the West. If not, I'd say they have a solid puncher's chance of knocking the Lakers off their Western Conference throne.

How do we begin to make sense of adjusted plus-minus?
JK: Outside of the obvious conclusion, which is "no one stat or metric, no matter how advanced or intricate, is ever going to come close to saying everything about one player," I have two thoughts on adjusted plus-minus.

The first is that I get how the basic +/- you see in box scores and 82games.com's version of plus-minus work, but I still don't totally understand how advanced plus-minus works, and that's a problem. I mean, I get the theory, that it adjusts for having good or bad teammates or playing against good and bad opponents, but how exactly does it define "good" and "bad"? Is "good" based on the other guy's adjusted plus-minus, or is the value of others derived from something like Player Efficiency Rating? Aren't both approaches problematic? Right now, adjusted plus-minus is sort of "He's good. Trust me," which I have trouble swallowing as a fan and certainly can't use to convince friends or readers of a guy's value.

The second problem is one that will get fixed over time, which is that we still don't really know how to read plus-minus type stats yet. We know with a stat like field goal percentage that a shooting guard is going to have a lower field goal percentage than a center, but we also know that the guard is probably shooting more 3s, shooting his free throws better and taking tougher shots than the center. We know how to read that stat.

But because plus-minus is one number and so nebulous, we don't know which plus-minus numbers to take with a grain of salt and which ones not to. I'll bring up the semi-infamous Durant example here. Durant had terrible +/- ratings for his first two seasons, but has been incredible in year three. Was the Durant phenomenon ever even real, or did Durant actually improve this year in ways the stats didn't see? If we want plus-minus metrics to be as legitimate as the box score ones, we have to stress-test it like we have the conventional numbers that came before them.

KA: I'm drawn to adjusted plus-minus because I'm desperate to find any metric that will approximate a player's defensive value, something we just don't have the tools to do right now. I'm more faithful than I probably should be given the lack of stress tests you talk about. Your point is well-taken and I'd add that stats like these are only valuable to the extent that they're predictive. There will always be players who make colossal jumps or experience unusual crashes in productivity, but apart from outliers, a stat must be dependable enough to offer a clear -- if general -- estimation of what that player is worth in the past, present and likely future. I've begun to spend more time examining the adjusted plus-minus numbers of 5-man units rather than individuals, in part because it seems more practical.

I suspect we'll know a lot more in three to five years than we do now. The metric's practitioners (and the people who trust them) will have a better sense of where the numbers skews, what those number might miss and the kind of noise those numbers create. In the meantime, I'll continue to watch the 2-year figures (and eventually 3-year, and 4-year). Any system that values Dwyane Wade, LeBron James, Chris Paul, Steve Nash and Kobe Bryant as the five best players in the NBA has to be on to something, right?

• How do you manage only 64 points against Toronto? Detroit's effort last night was painfully familiar to Dan Feldman of PistonPowered. Instead of composing a fresh recap, all he had to do was dust off the one from the previous game against Charlotte: "A few small fixes, and last night’s works just fine."
• John Krolik of Cavs the Blog took some strong impressions away from watching Tyreke Evans last night. The Cavs couldn't find anyone the guard the Kings' rookie over the first three quarters. Then Cleveland remembered it has this very athletic, physically impressive, relatively young wing they can utilize in case of defensive emergency.
• The Kings were in position to win the game. The fact that they didn't wasn't a result of youthful inexperience, says Zach Harper of Cowbell Kingdom: "At a certain point, [the Kings] got caught up in trying to make Evans a star. If he beats LeBron one-on-one a couple of times, you might lead off SportsCenter and get every national blog linking to you for the fourth time in a week. You might convince people to keep coming to the games (great crowd at the game last night by the way). Take a lesson from what LeBron James did: find your teammates and make the smart play when the offense isn’t going your way."
• Wendell Maxey, writing for the Portland Sentinel, recounts a fun anecdote about how in 1986-87, then-rookie Nate McMillan earned the respect of his Portland teammates. The story involves Maurice Lucas, McMillan telling off a veteran and a Coke machine.
• Michael Schwartz of Valley of the Suns says that Phoenix lacks toughness. That's a general problem, of course, but particularly costly on the offensive glass: "For a team that thrives on playing with rhythm and speed, it is nearly impossible to win giving up 16 offensive rebounds. Not only is an offensive board and bucket a huge momentum-changer, it slows down the pace of the game, which is something the Suns obviously don’t want to do."
• It was a big win for the Thunder in Phoenix. Royce Young of Daily Thunder took note of a creative decision by Scott Brooks: "I found it interesting that Thabo [Sefolosha] guarded Steve Nash for most of the game. Maybe that’s because Nash likes to hide behind screeners and shoot or fall away and shoot and maybe Thabo’s length could bother him. Thabo didn’t have a point, but I thought he played Nash well, especially in the second half. After nine assists in the first half, Nash had just three in the second." That's another thing that Russell Westbrook gives you -- the ability to put your most physical perimeter defender, Sefolosha, on the ball at strategic moments. Because Westbrook is capable of cross-matching, Brooks has the luxury to experiment and switch things up defensively.
• The Clippers will have a decision to make with Marcus Camby who, at 35 years old, is having one of the most productive seasons on his career. Camby is on the final year of his current deal and the Clippers have Blake Griffin coming back at some point in late January. D.J. Foster debunks the idea that if you don't trade a player with an expiring contract before the deadline that you're "letting him walk for nothing." Cap relief is often far more valuable than whatever a team might acquire in return for such a deal.
• The Kamenetzky Brothers always deliver with their podcast. When the Cleveland Plain Dealer's Brian Windhorst is the featured guest, it's tremendous stuff.
• Clark Matthews of the Daily Thunder reviews some of Oklahoma City general manager Sam Presti's greatest hits.
• According to Jeremy Schmidt of Bucksketball, the Wizards deployed an "attack Brandon Jennings" strategy to beat the Bucks at the Bradley Center Wednesday night. Jennings gambles too frequently, gets taken out of plays by strong screens, and bigger guards have been backing him down effectively. But rookie point guards almost always confront steep learning curves defensively, and Jennings will improve.
• For teams like the Trail Blazers in need of a big body, Scott Schroeder of Ridiculous Upside runs down the list of front-court candidates the D-League has to offer. In the backcourt division, D-League Digest's Steve Weinman is a fan of four-year Memphis guard Antonio Anderson, who's playing for Rockets affiliate Rio Grande Valley. While we're on the topic of RGV's guard rotation, I still believe Garrett Temple will play meaningful minutes in the NBA at some point.

Forget about Kobe -- the Nuggets have their hands full with the Lakers' seven-footers. Orlando needed to learn how to win -- it took them all of 72 hours. And Rick Sund deserves an "Atta Boy," in Atlanta.  

Jeremy Wagner of Roundball Mining Company: "You can talk about Kobe Bryant all you want, the Nuggets biggest concern should be how they can handle [Andrew] Bynum and [Pau] Gasol. With the starters on the floor Kenyon Martin is going to have to guard one of them and he has a serious length disadvantage against both. Most likely Kenyon will be guarding Gasol and for all his defensive desire and talents he is in a big hole ... Pau can shoot his 15-18 foot set shot over Kenyon at will and when he goes into the post his jump hook will be impossible for Kenyon to stop. Nene is relatively better equipped to cover Bynum than Kenyon is for guarding Gasol, but Bynum still has a significant length and weight advantage over Nene. On the other hand, Nene has done a decent job against Gasol in the past so will Denver choose to stick Kenyon on Bynum and double the heck out of him should he get the ball in the post thus creating one major mismatch instead of two less than desirable matchups?"

Zach Lowe of Celtics Hub: "This Celtics team could not play championship-caliber defense consistently. Their defensive numbers slipped a bit against Chicago, a mediocre offensive team, and it was likely, if not inevitable, that Orlando was going to score on Boston at least once or twice in this series. And the Celtics could not rely on their offense and their three-point shooting to carry them, as they did against the Bulls. Orlando's defense was the best in the NBA this season by some metrics. The tendency will be to look for what the Celtics did wrong -- to ask why Doc Rivers waited so long to try a small lineup, to wonder why Ray Allen shot so poorly until Game 7, to ask why the Celtics defenders had so much trouble guarding Mickael Pietrus tonight, why Eddie House couldn't get free, and on and on and on. The reality is that Orlando is a very good basketball team that presents major match-up problems for Boston sans [Kevin] Garnett."

Bret LaGree of Hoopinion: "Considering both the ownership situation and his brief tenure I'm going to give Rick Sund the benefit of the doubt for the time being. I don't believe that's simply a matter of being fair. By signing Flip Murray and Maurice Evans for a combined $4 million he earned the benefit of the doubt. The veteran pair combined to ably back up three positions while helping to accelerate the team's transformation, one which started following last season's trade for Mike Bibby, from an offense incapable of making three-point shots into a more diverse and dangerous team to guard."

THE FINAL WORD
Cavs the Blog: John Krolik revisits the Cavs-Magic regular season matchups.
Orlando Magic Daily: Much respect for Mr. Paul Pierce.
Daily Thunder: Thabo Sefolosha -- the next Shane Battier?

(Photos by Noah Graham, Brian Babineau, Scott Cunningham/NBAE via Getty Images)

The Hornets prevail in a must-win game over the Spurs, in which Bruce Bowen records a DNP-CD. Stephon Marbury is starting to figure out his role with the Celtics, while Iverson will have to adjust to his in Detroit. Read all about "Sixth Men: Past, Present, and Future" at the TrueHoop Network: 

Ryan Schwan of Hornets247: "Simply put, Chris Paul came out at half time and proved he was the best player on the floor.  I could fill up an entire observations section just with all the incredible plays he pulled out in that game.  It's such a joy to watch him play. As what usually happens in good wins with the Hornets, [David] West carried the team in the first half, scoring 14 and serving as the focal point for the offense.  In the second, Paul shifted from fourth gear to Warp 9 and carried the team to victory ... That was a big game, and it went into the 'Do Not Delete' section of my TIVO, so when I am without a game to watch in the off-season, I can fire that one up.  Winning without Peja, Tyson and Posey was pretty big."

Timothy Varner of 48 Minutes of Hell: "As Coach Popovich creeps closer to setting a rotation, it appears that Ime Udoka will get minutes behind Michael Finley. I'll stop short of making bigger pronouncements. It was only one game. Popovich is certain to use [Bruce] Bowen as a spot defender between now and the time he retires.  But I have to say, Pop is taking a gamble. Udoka is a tough-nosed defender, but even at his best moments he is not a versatile, game changing defender like Bruce Bowen. Bowen is a special player in that way. Or, reading into Pop's decision, Bowen was a special player in that way.  But Udoka does do some things better than Bowen. His offense is more varied (and erratic), he can handle the ball, and his rebound rate is 10.6, making him one of the better rebounding small forwards in the league. Defensively, Udoka does a better job against balky players like Ron Artest. But unlike Bowen, Pop won't call his number against Chris Paul -- he'll put George Hill into the game. If Sunday's rotation more or less sticks for the postseason, Popovich's gutsy decision to favor Udoka over Bowen will play a prominent role in determining San Antonio's championship aspirations, for good or ill."

Brian Robb of Celtics Hub: "Starbury only scored 2 points on 1/4 shooting but he did have 7 assists compared to just 1 turnover in 22 minutes to go with a +12 on the floor. There have been some growing pains in the past 10 plus games for the point guard but he is finally starting to look comfortable with the bench unit by distributing the ball to his teammates in the right spots ... a lot of these assists came off of some nice penetration, allowing him to draw multiple defenders to create dunks and open jumpers for his teammates. Great news to see him putting it together at the right time."

THE FINAL WORD
Piston Powered: Allen Iverson, Sixth Man -- A History.
Daily Thunder: Are OKC's best players named Sefolosha and Weaver? 
Raptors Republic: Toronto is putting all the pieces together ... in late March.  

(Photos by Layne Murdoch, D. Lippitt/Einstein, Joe Murphy/NBAE via Getty Images)

Watching the Thunder without Durant has been a revelation. Watching Paul Pierce is crunch time something to behold.  Just don't try watching a Celtics game in a Philly sports bar.  Watch and learn at the TrueHoop Network. 

Royce Young of Daily Thunder: "I don't deny that something's changed without Durant in the lineup. Some very, very good things have happened. The team has worked together and focused on all the details and little things. They knew in order to win, there couldn't be any room for error. There would be no 'Pass it to Durant, stand back and watch.' It was five working as one on both ends of the floor. Guys like Kyle Weaver and Nenad Krstic picked up their offensive games. Russell Westbrook was making everybody better around him. Thabo Sefolosha was a bear defensively and a smooth operator offensively. And when Durant steps back on the floor ... things should be better for him because he'll have a group of guys playing with more confidence and a better understand of their role...

One thing to remember is that Durant hasn't really has the chance to play with Thabo yet. KD hasn't been able to reap the benefits of a teammate that can wreak havoc and entirely change a game by himself on the defensive end. Thabo can pull off and help with the best of them. And he's also a better offensive player than people give him credit for. He plays smart, takes good shots, makes solid cuts, moves well without the ball and is a solid passer ... So Durant and Thabo really haven't had a ton of burn together. And Thabo is clearly a very important key to winning some of these games...

Playing without Durant has been a great thing for the team. Everybody has had to elevate themselves and play better ... Hopefully with this solid stretch without him, the team has realized that they are a major part of this success. It's not Kevin Durant and four other guys. It's five guys and one of them happens to be Kevin Durant."

Zach Lowe of Celtics Hub: "Paul Pierce has taken ONE HUNDRED freaking free throws in crunch time. 100!

... That is a remarkable number. Pierce has shot 445 free throws this year, meaning a full 22 percent of his free throw attempts have come in 30 fourth quarters and three overtimes–a span that makes up just 12 percent of the season so far. Read another way, Pierce has gotten to the line twice as often in 'clutch' fourth quarters as his normal FTA rate suggests he should have. Pierce has taken 26.7 percent of the Celtics 1,666 free throws overall this season; he's taken 37.5 percent of the free throws in my 'clutch' sample.

There's one obvious caveat here: About one-quarter (maybe a little bit less) of these 'clutch' foul shots came in end-of-game scenarios where the other team fouled Pierce to stop the clock. Even with this in mind, the number of FTAs is still enormous. And it's a hugely important skill. Jump shots go hot and cold, ingenious plays break down, but Pierce's will to attack the rim never wavers. Pierce at the foul line is probably the C's best late-game weapon -- even better than a Pierce jumper.

The Celtics also rely heavily on that. Pierce's 97 total two-point 'clutch' FGAs make up 22.5 percent of the Celtics 'clutch' total. Overall this season, Pierce has taken 17.7 percent of the team's two-point field-goal attempts. Clearly, the team leans on the Truth when games get close."

Rob Mahoney of Hardwood Paroxysm: "[Don Nelson] is the 'mad scientist.' the offensive wizard, and the cooky, unconventional coach who escapes to Maui in the offseason.  His coaching style appears whimsical, and so he appears whimsical.  But now, more than ever, we need to realize that he, as a person, is not.  His basketball mind is a unique one, and he's done plenty of things that make me smile and think about the game in new ways.  But time and time again, he has angered everyone around him, and split town with more enemies than friends.  Mullin had a good gig going before The Whimsical One waltzed into town, and though his performance was poor, his job never seemed to be in jeopardy.  One Don Nelson later, Mullin seems to be the one packing his bags while Nellie coasts through the end of the season, laughing maniacally on that extended flight to the islands.  Nellie might be doing us all a favor by getting rid of Mullin, but is a good move for all the wrong reasons still a good move?"

THE FINAL WORD
The Painted Area: A smart assortment of basketball book notes.
Hornets247: Tyson Chandler is coming along.
Celtics Hub: "This is Philly. You're not getting the [flippin'] Celtics game on in here.”

(Photos by Layne Murdoch, Garrett Ellwood, Frederic J. Brown/NBAE via Getty Images)